HYDROCARBONS

PROCESSING
A PRIMER FOR ALASKANS

by Arion R. Tussing and Lois S. Kr~m-er -J

ltiSTITUTE OF SOCIAL A~DU~z~;.~:;i; ~~i~~~~ . \

ISBN No. 0-88353-030-9
ISER Report Series Number: 53

Published by:
Institute of Social and Economic Research
UniversitY of Alaska
Lee Gorsuch, Director
707 "A" Street, Suite 206
Anchorage, Alaska 99501
1981

Printed in the U.S.A.

. , .. <"'7"'
 HYOROCARBONS
PROCESSING
.
'

AN INTRODUCTION TO PETROLEUM REFINING

AND PETROCHEMICALS FOR ALASKANS

ARLON R. TUSSING
LOIS S. KRAMER

J
Prepared for the Alaska State Legislature
with support from the Joint Gas Pipeline Committee
(Alaska State Legislature) and The Ford Foundation
August 1981

INSTITUTE OF SOCIAL AND ECONOMIC RESEARCH

- - - - - - - - - ---- - - ----

UN IVE RS ITY OF ALASKA

'
 PREFACE

This primer on fuels refining and petrochemicals man-

ufacturing is intended to illuminate a continuing debate
among Alaskans over the economics and potential benefits
of processing hydrocarbons in the state. The recurring
policy issue toward which the report was initially directed
was the question of what to do with the State's royalty
share of the crude oil, natural gas and natural-gas liquids
(NGL's) produced at Prudhoe Bay. The uses and implications
of our report will extend far beyond decisions regarding the
disposition of State royalty hydrocarbons, but this issue
provides a convenient focus for reviewing:
How the refining and petrochemical industries
are organized.
The role of petroleum and the petroleum indus-
try in Alaska.
Elementary hydrocarbons chemistry.
Refining and petrochemical-manufacturing tech-
nology.
Health and safety issues.
The economics of hydrocarbons processing.
The outlook for refining and petrochemicals in-
vestment in Alaska.
The first stage in the debate over disposition of State
hydrocarbons royalties ended in l 977, when the Alaska
Legislature conditionally approved the sale of the State's
Prudhoe Bay royalty gas to subsidiaries of the El Paso
Company, Tenneco, and Southern Natural Gas Company,
hoping that the political influence of these firms would lead
the federal government to select an "all-Alaska" pipeline
route for the Alaska Natural Gas Transportation System
(ANGTS). Under this plan, a plant producing liquefied
natural gas (LNG) and possibly other gas-processing facili-
ties would have been built at the pipeline's Gulf of Alaska
terminal.
iv Preface

The royalty-gas sales contracts lapsed later in

1977, when the President and Congress and the Canadian
government chose the Alaska Highway (" Alcan") pipeline,
sponsored by Northwest Energy Company and the Foothills
group, over El Paso Alaska's proposed LNG system and the
Mackenzie Valley pipeline proposal advanced by the Arctic
Gas group.
The second stage of the debate opened in 1978,
when the Legislature considered a long-term contract to sell
85 percent of the State's North Slope royalty oil, up to 150
thousand barrels per day (mb/d), to the Alaska Petro-
chemical Company ("Alpetco") if the company built a
"world-scale" petrochemicals plant in Alaska.
The Alpetco contract was later amended to
permit the sponsors to build a 100 mb/d fuels refinery,
which might or might not have produced petrochemicals.
Several changes in the project's ownership structure led to
its final sponsorship by the Alaska Oil Company, a subsidiary
of the Charter Oil Company. In May of 1981, Charter
abandoned its plan for a refinery at Valdez, stating that it
had been unable to obtain outside financing, and gave up its
right to purchase 7 5 mb/d of State royalty oil prior to
completion of the refinery.
Most recently, in 1980, the Alaska Department
of Natural Resources (DNR) entered into an agreement with
the Dow Chemical Company, the Shell Oil Company, and a
group of associated companies to study the feasibility of
transporting and processing Prudhoe Bay NGL's in Alaska.
The State, in turn, granted the participating companies an
option to buy its royalty share of the NGL's and agreed to
use its influence with the North Slope producing companies
to obtain additional feedstocks for the petrochemicals com-
plex, should it prove feasible. This study is scheduled for
completion and delivery to the State in September, 1981.
Concurrently with the Dow-Shell study, the Ex-
xon Chemical Company has independently been studying the
feasibility of petrochemicals manufacturing based on North
Slope NGL's. Earlier in 1981, Arco made its own assessment
Preface v
of Prudhoe Bay methanol production as an alternative to
construction of a natural-gas pipeline.
The several projects have been quite different techni-
cally, but all of them tend to evoke similar hopes, fears, and
controversy among Alaskans. The hopes and arguments
favoring such ventures have been: increased local "value-
added" from the state's natural resources (as opposed to
their export in unprocessed form), the contribution that this
processing would make to the state's economic growth and
economic diversity, a greater and more diversified tax base,
new and more diverse job opportunities, and lower Alaska
prices for fuels and other petroleum products.
At the same time, some Alaskans have been skeptical
about the underlying economic soundness of the proposals
and feared that one or more of these ventures might
ultimately have to be rescued by the State treasury. Other
concerns have been the possibility that long-term royalty-
gas export contracts could foreclose future opportunities for
residential, commercial, industrial or electric-utility use of
the gas in Alaska, and that long-term royalty-oil sales to
export-oriented new refineries could leave existing refiner-
ies that serve Alaska customers short of raw material, if the
decline in Prudhoe Bay production made the oil producers
less willing to sell crude oil to these refineries.
Other potentially adverse impacts are the prospects of
deepening Alaska's already excessive dependence on petro-
leum-related industry, and of once more repeating the
state's familiar boom-bust cycle; new sources of pollution
and other health, safety, or aesthetic hazards; and un-
welcome changes in community values and life-styles.
To aid the rational discussion of such issues, this
primer tries to set in context the basic technical and
economic facts, analytical concepts, and policy considera-
tions relevant to hydrocarbons processing in Alaska. Many
of the crucial questions have already found their way into
public debate and set the stage for our discussion of the
more technical aspects of fuels refining and petrochemicals
processing. These questions, for example, include consider-
ations of:
vi Preface

Feasibility. Is Alaska a realistic location

for nationally or internationally competitive
fuels refining or petrochemicals-manufacturing
activity?
Type of industry. For what specific kinds
of hydrocarbons-processing, if any, does Alaska
have a special comparative advantage, and what
kinds of facilities are especially unpromising for
Alaska?
Interrelationships. What interrelationships
exist among the projects that have been pro-
posed? Are some of them mutually exclusive?
How will decisions regarding ANGTS affect the
viability of a gas-liquids pipeline or natural-gas
liquids-based petrochemicals production, and
vice-versa?
Influence of the State. What special abil-
ity does the State's ownership of royalty oil and
gas, regulatory powers, taxing authority, or in-
vestment capability give it to encourage or dis-
courage investment or to affect the character or
location of facilities that process Alaska hydro-
carbons? And, to what extent is it(!) proper or
prudent in a society committed to private enter-
prise, or (2) in the interests of Alaskans, that the
State government deliberately use its powers to
influence the course of development?
Direct economic impacts. How many jobs,
of what character, will each proposed project
offer in its construction and operational phase
respectively, and who will fill these jobs? How
will construction and operation of the facilities
affect the demand for services in other local
industries?
Indirect development impact. To what
extent will the existence of any of the projects
in question stimulate (or discourage) investment
in complementary (or competing) industries, and
Preface vii

what will be their total impact on the state's

economy after taking into account all their short
and long-term, direct, indirect, and multiplier
effects?
Health, safety, environmental, and aesthe-
tic considerations. To what extent do the propo-
sed projects (or their indirect developmental
effects) have unavoidable adverse impacts or
create known or potential risks of adverse im-
pacts on health, safety, the natural environment,
or other dimensions of the quality of life in
Alaska?
Conflicting objectives. To what extent do
specific kinds of State efforts to attract refinery
or petrochemical investments assist or conflict
with other goals, such as maximization of royal-
ty and tax revenue from oil and gas production,
early completion of the natural-gas pipeline, or
availability of low-cost energy for local residen-
tial, commercial, or industrial consumption?
Consequences. What are the likely conse-
quences of making an early commitment or not
making such a commitment of the State's Prud-
hoe Bay royalty gas and/or gas liquids? Are
there additional costs that State and local go-
vernments may incur as a result of their aggres-
sive pursuit of petrochemical investment in
Alaska?
These questions, while not exhaustive, are the major
issues in the current public debate over State policies
toward petrochemical development. Although the authors
have tried to give general answers to some of these ques-
tions, the main function of the present paper is to provide
its readers with some of the background necessary to
develop their own answers.
viii Preface

Sponsorship. Production of this study was supported

by a contract between the University of Alaska's Institute of
Social and Economic Research and the Alaska Legislature's
Joint Gas Pipeline Committee with a matching grant from
the Ford Foundation. The contract with the Legislature was
administered by Susan Brody of the House Research Agency.
Acknowledgements. The authors are particularly in-
debted to officers and personnel of Exxon U.S.A., the Dow
Chemical Company, Shell Oil Company, Westcoast Trans-
mission Company, and Nova Ltd. and its subsidiary Alberta
Gas Ethylene Company for their assistance during the early
phases of manuscript preparation.
A review draft of this report was distributed in several
hundred copies at the end of June, 1981. This draft was
reviewed in detail at Exxon, Dow, Shell, in several units of
the Atlantic Richfield Company, and at Sohio Petroleum
Company. We cannot make all the necessary individual
acknowledgements here, because the persons involved in
these reviews were so numerous, and in many cases the
specific names are unknown to us. Special mention is due,
however, to Mark Wittow of the Legislature's Joint Gas
Pipeline Committee; Mary Halloran of the Alaska Depart-
ment of Natural Resources; Robert Mohn of the Alaska
Power Authority; Robert Loeffler, Alaska's Washington
counsel; and Phillip Essley of the Alaska Project Office of
the Federal Energy Regulatory Commission (FERC). Connie
C. Barlow, vice-president of ARTA, Inc., reviewed the
substance, format, and expression of the draft in great
detail for the House Research Agency.
The authors accepted many but not all of the sugges-
tions and criticisms offered by these reviewers, and take
full responsibility for any errors or deficiencies in this final
version.
 TABLE OF CONTENTS

Chapter l Overview of Petroleum and the

Petroleum Industry ..... .......... l

Chapter 2 The Petroleum Industry in Alaska . . . . . . 11

Hydrocarbon Resources,
Reserves, and Production . 11
Hydrocarbons Processing in
Alaska . . . . 16
Prudhoe Bay Natural-Gas
Reserves and ANGTS . . . . 19

Chapter 3 Fundamentals of Hydrocarbons

Chemistry .......... ......... 25
General Introduction . 25
Composition of Natural
Hydrocarbons . . . . . 25
Chemistry . . . . . . . 28

Chapter lf Fuels Refining . . . . . . . . . . . . . . . . . . . . 35

Petroleum-Industry Structure . 35
Feedstocks and Petroleum
Products . . . . . . 37
Refinery Products . lf l
Refining of Petroleum . . . . . . 51
Refinery Technology &
Design . . . . . . . 57
Forces for Change . 60
Outlook for the l 980's . . 66

Chapter 5 Petrochemicals . . . . . . . . . . . . . . . . . . . 69
Introduction . . . . . . 69
Chemical Industry Structure . 69
Petrochemical Feedstocks . . . . 73
Petrochemical-Product
Groups . . . . . . 77
Final Products . . . . . . . 81
x Contents

Petrochemical Processes and

Plant Design . . . 84
Petrochemical Complexes and
Utility Requrements , , . , . , , , , 91
Health and Safety Issues 96
Options for Regulation . . . 102
Prescriptive vs. Economic
Remedies ~ . . . . . . . . . . . . 103

Chapter 6 The Economics of Hydrocarbons-

Processing and the Outlook for Re-
fining and Petrochemicals in
Alaska . . . . . . . . . . . . . . . . . . . . . . . . . ll l
Hydrocarbons-Transportation
Economics . . . . 111
Fixed-Capital Costs . . . ll6
Feedstock Costs and Supply ll 9
Illustrations . . .. . . 121
Economies of Scale 131
Analyzing Project Feasibility 134
Coping with Uncertainty and
Risk . .. 139
 CHAPTER 1
OVERVIEW OF PETROLEUM
AND THE PETROLEUM INDUSTRY

Crude oil, natural gas, and natural-gas liquids are all

"petroleum," which is the general term for hydrocarbons ---
compounds found in the earth's crust that are composed
mainly of hydrogen and carbon atoms. Hydrocarbons vary
considerably in molecular size and structure, and each
hydrocarbon compound can exist as a solid, a liquid, or a
gas, depending on the pressure and temperature to which it
is subjected.
"Crude-oil" fields or reservoirs are naturally occurring
deposits containing hydrocarbons that are liquid at atmo-
spheric pressures and temperatures, while "natural-gas"
fields or reservoirs are deposits containing only hydro-
carbons that are gases under the same conditions. However,
most commercially recoverable petroleum deposits, includ-
ing the Prudhoe Bay field, contain a mixture of liquid and
gaseous hydrocarbons that have to be separated in the field
for transportation and processing.
The petroleum industry, as we define it for the pur-
poses of this primer, includes businesses engaged in finding
and extracting hydrocarbons from the earth, and their
storage and transportation; the refining, distribution, and
sale of fuels and lubricants; and related service and support
activities. It also includes a "petrochemical" sector --- the
manufacture and distribution of organic chemicals based
upon petroleum feedstocks, often by affiliates of petroleum
production and refining companies.
Internationalism. Petroleum is the most important
commodity in world trade in both volume and value, and a
large part of the world's total production of petroleum
liquids is transported and processed by a few multinational
companies. The reason for the industry's exceptional inter-
nationalism is the widely differing locations of the chief
petroleum-producing areas and the major markets for petro-
leum products. Figure 1 illustrates the geographic disparity
between global oil production and consumption in 1979.
Page 2 Overview

Figure 1
Crude-Oil Production and
Consumption of Refined Products By Area, 1979

Production Consumption
(l ,000 barrels) (l ,000 barrels)

United States 3' 111, 625 6,728,410

Canada 545,675 691,675
Latin America 1,912,200 1,604,175
Middle East 7,803,700 542,025
Africa 2,401,700 478,150
Asia/Pacific 1,042,075 3,429, 175
Western Europe 826,725 5,427,550
Communist Nations 5,120,950 4,688,425

Source: British Petroleum Company, Basic Petroleum

Data Book, 1980.

Size and capital intensiveness. The world petroleum

industry is both very large, and as a whole, exceptionally
capital-intensive. Six of the ten largest firms in the 1980
Fortune 500 were oil companies. Table 1 summarizes the
Overview Page 3

Chase Manhatta~ Bank's survey of the petroleum industry's

197 5-77 capital expenditures. The industry's new invest-
ments over the three years amounted to about $62 billion in
the United States and $168 billion worldwide.
Table l
Domestic and Foreign Capital Expenditures of
The World Petroleum Industry
Production Transeortation
U.S. Forei n Total U.S. Foreign Total
$bi! % bi! % $bi! $bi! % $bi! % $bi!
1975 9.4 49 9.9 51 19.3 3.7 25 11. 5 76 15.2
1976 13.4 52 12.3 48 25.8 3.9 24 12.4 76 15.2
1977 15.2 47 17.3 53 32.5 2.3 22 8.1 78 10.4
Refineries & Chem. Plants
U.S. Total U.S. Total
$bi! % $bi! $bi! % $bi!
1975 3.6 30 8.3 70 11. 9 .6 30 1. 5 70 2.2
1976 3.8 33 7.6 67 11.4 .6 29 l. 6 71 2.2
1977 3.7 25 11.0 75 14.7 .8 29 1.9 71 2.7
Other Ca[>ital Seending Total Capital Spending
U.S. F orei n Total U.S. Forei n Total
$ bi! % bi! % $ bi! $ bi! % bi! % $ bi!
1975 .4 34 .7 67 1.1 17.7 36 31.9 64 49.6
1976 .3 29 .8 71 1.1 22.l 39 34.6 61 56.7
1977 .4 31 1.0 69 1.4 22.4 36 39.2 64 61.2

Source: The Chase Manhattan Bank, Capital Investments

of the World Petroleum Industry.

Table 2 shows that petroleum refining also had the

highest ratio of assets per employee among the 29 indus-
tries included in the For tune survey. The chemical industry
ranked sixth. Petroleum was also in first place among all
industries with respect to the ratio of assets to sales.
Not all phases of the industry are exceptionally capi-
tal-intensive, however. In the Middle East, for example,
Page 4 Overview

capital costs for crude-oil production --- the cost of wells,

gathering lines, and separating facilities --- tend to be
relatively low, ranging from about $100 to $500 for the
capacity required to produce one barrel of oil per day. At
an oil price of $32 per barrel, even $.500 per "daily barrel"
implies that only 16 days of production would be needed to
recover the fixed investment in field development.

Table 2
Assets per Employee for the Fortune 500
(Industry Medians)
Petroleum Food 49,ti88
refining $303,839 Motor Vehicles ti6,039
Mining, crude- Shipbuilding, rail-
oil production 254,336 road & transport
Broadcasting, equipment ti3,9til
motion-picture Rubber and plastic
production and products ti2,563
distribution 108,772 Measuring, sci en-
Tobacco 81,937 tific and photo-
Metal manufact'g 79,868 graphic equipm't ti! ,000
Chemicals 77,947 Aerospace ti0,901
Paper, fiber and Musical instrum'ts,
wood products 76,ltil toys, sporting
Pharmaceuticals 66,5ti3 goods 37,666
Publishing, Electronic appl. 37,59ti
printing 56' 129 Textiles & vinyl
Glass, concrete flooring 26,ti31
abrasives, and Apparel 20,36ti
gypsum 55,668 Leather n. a.
Industrial and Furniture n.a.
farm equipm't 53,361 Jewelry, silver-
Office equipm't, ware n.a.
computers ti9,039
All industries $ 55,505

Source: Fortune, May ti, 1981

Overview Page 5

The capital cost per daily barrel for new crude-oil

production in the North Sea, the United States Outer
Continental Shelf (OCS), or the Arctic is typically much
higher than for Middle Eastern production --- in the range
of $5,000 to $2.5,000 per daily barrel. Synthetic oil and gas
plants are expected to be even more capital-intensive, with
capital costs of $50,000 or more for the producing capacity
equivalent to one barrel of oil per day.
Refining and petrochemical plants also require very
large capital additions, both absolutely and per unit of
capacity: a completely new ("grass-roots") state-of-the art
oil refinery may cost more than a billion dollars --- at
$5,000 to $10,000 or more per daily barrel of capacity ---
and a first-stage petrochemicals plant may cost even more.
Even the extra equipment an existing refinery would need to
process lower-quality ("heavy" or high-sulfur) types of crude
oil tends to add new capital costs of $1,500 to $2,500 or
more per daily barrel of capacity.
High technology. The search for natural hydrocarbons
is reaching out to more remote and difficult locations:
farther below the earth's surface, under deeper water, and
into the Arctic. Construction of production platforms in the
North Sea marked the first time engineers had installed
permanent structures of any kind in such deep or wave-
stressed waters, while the Trans-Alaska oil pipeline (TAPS)
required radically new pipeline design and construction
techniques to cope with tundra and permafrost conditions.
The production of unconventional hydrocarbons or
even familiar resources in new environments (tar sands and
oil shales, for example, or natural gas in coal seams, tight
rocks, and pressurized brine solutions) is tied directly into
technological advance as is progress in the refining and
petrochemical sectors, where new end-products appear fre-
quently and where both the demand mix and feedstock mix
continue to change.
Short- and long-term flexibility. Petroleum-product
demand changes constantly. Part of this change is short-
term, determined by the seasons and the weather, or by
Page 6 Overview

economic conditions. Gasoline consumption peaks in the

summer, and heating oil consumption in the winter; a
warmer-than-usual spring increases gasoline demand, while
a colder-than-usual winter favors heating-oil demand. Con-
sumption of all petroleum products and petrochemicals
tends to fall off in recessions, though not in constant
proportions. These fluctuations require that refiners be able
to vary the mix of different products in their plant output,
carry some surplus processing capacity, and maintain stor-
age facilities.
Another part of the change in demand is longer-term.
It appears that the total consumption of petroleum products
in the United States and the world as a whole peaked in
1978, and the subsequent decline may well be permanent.
Gasoline and residual oil demand, in particular, are expected
to continue shrinking, but the consumption of "middle distil-
lates" (diesel fuel, home-heating oil, and jet fuel) and the
use of refinery products for petrochemical feedstocks may
resume their growth or at least stabilize. At the same time
that gasoline consumption as a whole is shrinking, the U.S.
Environmental Protection Agency (EPA) is requiring refiners
to phase out lead as a gasoline additive, compelling them to
produce an essential! y new kind of gasoline in order to
obtain acceptable anti-knock ("octane") ratings.
These shifts in product demand are all occurring at a
time when "heavy" crude oil containing a high proportion of
residual oil is becoming a relatively larger part of the total
oil supply. The continuing decline in Lower lf8 natural-gas
production from traditionally-exploited kinds of resources is
also reducing the supply of natural-gas liquids (NGL's), a
major feedstock source for the petrochemical industry. As
a result, an overall decline in the demand for petroleum-
processing capacity may not forestall the need for further
investment in facilities to "upgrade" surplus residual oil into
middle distillate fuels and petrochemical feedstocks like
naphtha and gas oil.
Overview Page 7

Long lead times. Capital-intensiveness and high tech-

nology imply long engineering lead times and long construc-
tion schedules, with heavy capital outlays required far in
advance of any return on investment. Refineries, petro-
chemical plants, frontier oil and gas development, and
pioneering pipeline ventures like TAPS and ANG TS, tend to
require 3 to 8 years or even longer for their planning,
design, construction, and shakedown.
Risk. Risk and uncertainty pervade all segments of
the petroleum business. Geological or exploration risk --the
low percentage of "wildcat" wells that lead to commercial
oil and gas discoveries --- probably receives the greatest
emphasis in public discussions of petroleum industry risks.
But the most significant risks in the industry today tend,
rather, to concern costs, markets, and political and regula-
tory treatment. The large absolute size of individual
projects and the long time that typically elapses between
the initial outlay and its return make the economics of new
refineries, processing plants, or transportation systems ex-
tremely sensitive to future raw materials costs, product
markets, tax treatment, and government policies for long
periods into the future. Such investments are therefore
exceptionally vulnerable to cost overruns, unforeseen chan-
ges in raw materials costs or supply interruptions, and to
changes in product demand, tax treatment, regulation and
other government policies.
Vertical integration. Vertical integration is primarily
an attempt to reduce the supply and market risks faced by
the various sectors of an industry. Primary raw materials
producers are likely to integrate "downstream" into refining,
chemical manufacturing, and distribution, in order to assure
themselves a long-term market. At the same time, refiners
and processors try to obtain "upstream" control over produc-
ing properties in order to stabilize their raw materials costs
and reduce the possiblity that expensive plants will become
idle or customers go unserved in some future feedstock
shortage. Crude-oil pipelines are typically built and oper-
ated by major producers and/or refiners, because only they
can assure that the pipeline will be used.
Page 8 Overview

As a result, a relatively small number of multinational

firms produce, transport, refine, and market most of the
petroleum liquids in the United States, but these major
companies share the stage with independent and partially-
integrated producers, refiners, resellers, and marketers of
all sizes.
The chemical industry is more concentrated than the
oil industry: the five top companies accounted for 60
percent of total U.S. chemical sales in 1979, in contrast to
the five top refiners' 48 percent of petroleum-product sales.
In the last five years, however, growing downstream inte-
gration by major oil companies has given them a dominant
role in production of primary petrochemicals, such as ethyl-
ene and benzene.
Government involvement. Governments powerfully
influence the structure and performance of the petroleum
industry through their roles as landlords and royalty-owners;
tax collectors; protectors of investors, consumers, and com-
petitors, and of health, safety and the environment; price
regulators and allocators; statisticians; traders; and promo-
ters or investors.
Some government programs or policies have encou-
raged vertical integration. Examples have been percentage
depletion allowances and the windfall profits tax, both of
which permit integrated companies to reduce their federal
tax burden by means of crude-oil "transfer prices" that shift
book profits to the stage of production where the tax rate is
lowest. Other programs have penalized integration: The
import quota system that existed between 1958 and 1974,
for instance, had a "sliding scale" that favored small refin-
ers, while the "small refiner bias" in the crude-oil "entitle-
ments system" under the Emergency Petroleum Allocation
Act had a similar effect between 1974 and 1980.
The government of Alaska is distinctive among the
states because of the size of the petroleum resource it
controls and particular! y because of the disproportion bet-
ween this resource base and the state's present population.
At the end of 1980, the State's royalty interest in just
Overview Page 9

proved reserves amounted to about 1.1 billion barrels of

crude oil and NGL's, and 3.9 trillion cubic feet (tcf) of
natural gas. Its taxing authority extended to another 8.6
billion barrels and 32.8 tcf of proved reserves. Further oil
and gas discoveries will surely add to these totals.
With a 1980 Alaska resident population of about 400
thousand persons, these supplies exceed by many times any
reasonably foreseeable demand by the State's existing resi-
dential, commercial, or industrial consumers. The expected
revenues from extracting these resources will likewise far
surpass the population's need for the ordinary services of
State and local governments, leaving a large current reve-
nue surplus available for long-term investments, industrial
development projects, or direct distribution.
Thus, Alaska's discretionary powers over the oil and
gas itself, and over the revenues they generate, are excep-
tional. The role of State government as resource owner,
manager, regulator, and potential investor plunges the issues
of refinery and petrochemical development squarely into the
political arena. As Alaska's oil and gas industry is already a
quarter-century old, a brief overview of its existing and
contemplated developments will shed some light on how and
where the industry may develop in the future.

 CHAPTER 2
THE PETROLEUM INDUSTRY IN ALASKA
The kind, size, and location of existing petroleum-
related activity in Alaska will doubtless have a large influ-
ence on the kind, size, and location of future refining and
petrochemical investments.
Hydrocarbon Resources, Reserves, and Production.
Cook Inlet. In the modern era, the first commercial
discovery of petroleum occurred in 19 57 at Swanson River
on the Kenai Peninsula, 100 kilometers southwest of Ancho-
rage. The last major oil discovery in the upper Cook Inlet
region was in 1965, and the last important gas discovery was
in 1966. Oil production peaked in 1970 at 229 thousand
barrels per day (mbpd), averaged 85 mbpd in 1980, and is
continuing to decline rapidly. Industry geologists believe it
is unlikely that new discoveries in the Upper Cook Inlet area
will reverse this trend.
Natural-gas production, other than volumes reinjected
to maintain oil-field pressures, averaged about 600 million
cubic feet (mm cf) per day in 1980. At the end of 1980,
Cook Inlet's proved natural-gas reserves totalled more than
3.5 trillion cubic feet, about 16 years' production at the
current rate. Exploration of the area is continuing to
produce promising "shows" of natural gas. Because about
half of the area's proved reserves are still not firmly
committed to production, however, there is little incentive
for the industry to develop new discoveries, and thus to add
them to the proved reserves category. In any event, given a
growing market, it is likely that Cook Inlet gas production
could continue to increase for, say, another decade before
beginning to fall off.
Prudhoe Bay area. The Prudhoe Bay oil and gas field
in Arctic Alaska, discovered in 1968, is relatively small
compared to a few fields in the Middle East (and perhaps in
the U.S.S.R.) but it is the largest crude-oil deposit yet
discovered in the United States or Canada, and one of the
Continent's three or four largest natural-gas deposits.
Page 12 The Petroleum Industry in Alaska

The main reservoir at Prudhoe Bay (the "Sadlerochit"

formation) began producing crude oil in commercial quanti-
ties when TAPS was completed in 1977. Current production
is at the reservoir's maximum allowable offtake of 1.5
million barrels per day, about 18 percent of the total U.S.
production of crude oil, and 15 percent of domestic petro-
leum liquids production.

Table 3
Alaska North Slope Crude Oil Production Forecasts, 1981-2000
(thousands of barrels per day; percentage confidence intervals)

low mean high low mean high

Confidence Level Confidence Level
Year (9 5%) (50%) (5%) Year (95%) (50%) (5%)
1981 1483 1500 1560 1991 759 1295 3270
1982 1483 1500 1560 1992 750 1289 3566
1983 1452 1558 1643 1993 720 1279 3541
1984 1452 1690 1815 1994 648 1185 3451
1985 1500 1771 1950 1995 584 1112 3429
1986 1530 1808 2133 1996 569 1094 3820
1987 1585 1906 2332 1997 521 1013 4055
1988 1410 1745 2440 1998 476 935 4220
1989 1092 1465 2798 1999 433 867 4020
1990 910 1377 2898 2000 394 791 3844

Source: Arion R. Tussing, "The Outlook for Alaska North

Slope Crude Oil Production: 1981-2000." Insti-
tute of Social and Economic Research, Research
Summary, January 1981.

Table 3 shows the range of likely crude-oil production

figures from all North Slope fields through the year 2000.
The "Kuparuk" and "Lisburne" formations in the Prudhoe Bay
area and other discoveries nearby, at Point Thomson-
Flaxman Island, Sag Delta-Duck Island, and Gwyrdyr Bay
will probably contribute an additional one or two hundred
thousand barrels per day by the time that production from
The Petroleum Industry in Alaska Page 13

the Sadlerochit reservoir begins to fall off in the mid-to-

late l 980's. Without very large additional discoveries,
however, there is little chance that production from new
fields on the North Slope will fully offset this decline.
Commercial production of natural gas from Prudhoe
Bay awaits completion of the Alaska Highway gas pipeline
("Alaska Natural Gas Transportation System" or ANGTS), no
sooner than 1986. Gas producers and pipeline sponsors are
counting on the Sadlerochit formation to produce at least
2.7 billion cubic feet (bcf) of raw, unprocessed gas per day,
the equivalent of about 2.0 bcf per day of pipeline-quality
gas, for 20 to 25 years. There are no authoritative public
estimates of potential production from the other known
reservoirs and recent discoveries in or around the Prudhoe
Bay field, but they might increase these figures by another
25 to 50 percent by the time the gas transportation system
is in place.
The outlook for further discoveries. Alaska and its
offshore margins contain the bulk of the remaining unex-
plored petroleum-producing prospects in the United States,
and major additional oil and gas discoveries are inevitable.
Some areas in and adjacent to Alaska are regarded as the
most promising acreage for petroleum exploration under the
American flag. Three examples of such Alaska areas are (1)
portions of the Beaufort Sea where the State and Federal
governments held an oil and gas lease sale in December
1979, (2) the St. George Shelf South of the Pribilof Islands
in the Bering Sea, and (3) the Arctic National Wildlife Range
(ANWR) in the extreme Northeast corner of Alaska.
On the basis of surface investigations and inferences
from drilling elsewhere, geologists believe that each of
these areas contains one or more geological structures
capable of containing a "supergiant" oil and/or gas reservoir.
(A supergiant is an oil field with recoverable crude oil
reserves of one billion barrels or more, and a supergiant gas
field is one with an equivalent amount of energy in the form
of natural gas --- roughly 1.8 tcf.)
Page 14- The Petroleum Industry in Alaska

Supergiant oil and gas discoveries are rare and random

events, however, and the probability that another field the
size of Prudhoe Bay will be discovered in this century is
very slim. Moreover, there is still no way short of drilling
to find out for sure whether even the most promising
structure identified from the surface contains petroleum
rather than, say, salt water.
The location and current status of these three explo-
ration prospects illustrate the problems of developing oil
and gas production in frontier areas of Alaska generally.
Many petroleum geologists consider the Beaufort Sea the
nation's most promising exploration frontier. The first
exploration well in the Beaufort Sea was "spudded" --- i.e.,
began drilling --- in November 1980, less than one year
after the 1979 lease sale, and at least one major oil
discovery has already been announced (in the Duck Island -
Sagavanirktok Delta area).
Nevertheless, it is not reasonable to expect any com-
mercial oil or gas production in less than 6 to 10 years.
Local villages, whaling interests, and environmental groups
have filed lawsuits against drilling on both State and Federal
acreage, and the resulting possibilities for delay are sub-
stantial. Even after all legal and regulatory obstacles are
overcome, the short shipping season, the horrible weather,
and the need to develop new engineering techniques for
finding oil and producing it from under ice-stressed seas will
contribute to further delays.
The area of the 1979 Beaufort Sea lease sale is under
shallow water within about a 200-kilometer radius of Prud-
hoe Bay and can rely to a large extent on the infrastructure
created to serve Prudhoe Bay --- particularly on TAPS and
ANGTS, should exploration be successful. However, the St.
George Shelf in the Bering Sea, where a Federal OCS lease
sale is scheduled for 1982, is far from the mainland. Any
exploration effort there must cope with much deeper water
and high waves as well as different but equally unhospitable
weather, And the petroleum industry has yet to establish
staging areas or even the beginnings of an oil or gas
The Petroleum Industry in Alaska Page 15

transportation system in the area. Finally, the State

government itself is on record opposing petroleum explora-
tion in the Bering Sea, along with some communities and
local interests, fishermen, and conservationists.
The ANWR is in the extreme Northeast corner of
Alaska. Its most promising acreage is on land or in shallow
seas immediately offshore, and although this area is con-
siderably farther from Prudhoe Bay, TAPS, and ANGTS than
the 1979 Beaufort Sea leases, exploration in the ANWR
would benefit considerably from existing North Slope infra-
structure development. It would, in addition, draw directly
on proved engineering techniques for Arctic tundra areas
developed for Prudhoe Bay operations.
No leases are yet scheduled for ANWR, however, and
preservation of the wilderness status of the Range is one of
the highest political priorities of national conservation orga-
nizations. As a result, the 1980 Alaska Lands Act closed
most of the ANWR to exploration, except for a strip along
the Arctic Coast, and even that parcel requires a 5-year
geological and geophysical study by the government, follow-
ed by Congressional action, before any leasing would be per-
mitted.
Every other prospective oil and gas exploration fron-
tier in Alaska differs somewhat from the three we have used
as illustrations, but almost all of them hold comparable
obstacles to development in the form of remoteness,
climate, novel engineering or environmental challenges, lack
of an existing infrastructure, and/or local, statewide or
national opposition.
In summary, therefore, Alaska doubtless has a great
deal of undiscovered petroleum, and petroleum exploration
will be an important activity in and offshore Alaska for
many decades. Apart from the known deposits in and
adjacent to the Prudhoe Bay field and in Upper Cook Inlet,
however, how much oil and gas will actually be discovered
and produced in the State, where, and when, are complete
mysteries. In no case can these unknown resources be a
basis for projecting the State's fiscal outlook, its future
Page 16 The Petroleum Industry in Alaska

population and economic activity, or future investment in

refining or petrochemical manufacturing.
Royalty oil and gas. Lease contracts covering the
established production at Prudhoe Bay and most of the Cook
Inlet fields reserve a one-eighth royalty interest in the oil
and gas produced, which the State may at its option take
either "in value" (cash) or "in kind" (as oil and gas). On
many State leases not yet under production, including the
Beaufort Sea lease area, the State's royalty share is one-
sixth or more. Taking royalty oil or gas in kind, in order to
sell it to an established or prospective in-state hydrocarbons
processor, has been and will likely continue to be one of the
State's tactics in attempting to encourage refining and
petrochemicals investment.
Outer Continental Shelf (OCS) oil and gas leasing is
under federal jurisdiction, however, and the State has nei-
ther a royalty share nor the right to levy taxes on OCS
production.
Hydrocarbons Processing in Alaska.
Estimates of 1980 Alaska petroleum-products con-
sumption by five different authorities range from 63 to 89
mb/d, of which about 28.5 mb/d appears to be jet fuel,
(much of it destined for international airlines and the
military, and thus not strictly an in-state use). The remain-
ing direct Alaska consumption of motor fuels, heating oil,
and electric utility fuel in Alaska was somewhere in the
range of 35 to 60 mb/d --- the most likely figure is on the
order of l/5 mb/d. Nearly half of this total was imported
from the Lower 1/8, much of it to Southeast and Western
Alaska.
Fuels refineries. The remainder of Alaska's petroleum
requirements are served by three in-state refineries, opera-
ted by the Standard Oil Company of California (Chevron) at
Swanson River on the Kenai Peninsula, by Tesoro Alaskan at
Nikiski (Kenai), and by Mapco (formerly Earth Resources
Company of Alaska) at North Pole near Fairbanks. Together
the three refineries have been running slightly more than
The Petroleum Industry in Alaska Page 17

JOO mb/d of crude oil and producing about 44 mb/d of

refined products, principally fuels. The balance of their
output is residual oil, which is shipped to the Lower-48 for
further processing or for sale as electric-utility fuel.
The Chevron refinery was built in 1963, has a crude-oil
distillation capacity of 22 mb/d, and refined an average of
13.5 mb/d in 1980. Chevron is currently considering shut-
ting-down this small and relatively inefficient plant because
of excess capacity in the company's larger West Coast
facilities. The Tesoro refinery was built in 1966 expressly to
run sweet (low-sulfur), light (high gasoline-content) crude
oil, which it obtains from the State in a long-term sale of
Cook Inlet royalty crude. Because of the decline in Cook
Inlet production, the refinery has been modified to run a
feedstock mixture that includes about 15 percent Prudhoe
Bay crude oil, which has a higher sulfur content and lower
"gravity" (less gasoline and more residual oil). The Tesoro
plant's crude-oil distillation capacity is 48.5 mb/d, and it ran
essentially at full capacity in 1980.
Both the Chevron and Tesoro refineries export about
half their total product to the U.S. West Coast --- mainly
residual oil and crude gasolines for blending - - - and sell
middle distillates (diesel, home heating oil, and jet fuel),
gasoline (Tesoro) and asphalt (Chevron) in Alaska.
The North Pole refinery is less complex than Tesoro's,
with a crude-oil processing capacity of 47 mb/d. In 1980 the
refinery processed an average of 43 mb/d of crude oil taken
from TAPS; the output consisted of 16 mb/d of middle
distillates sold in Alaska and 27 mb/d of residual oil, LPG's,
and crude gasoline reinjected into TAPS for processing by
Lower-48 refiners.
The refinery proposed for Valdez by the Alaska Oil
Company (Alpetco) to use Prudhoe Bay royalty oil would
have been much more sophisticated than the existing Alaska
plants in that it would have processed its entire crude-oil
input of 100 mb/d into light fuels (including high-octane
unleaded gasoline) and middle distillates. The refinery was
Page 18 The Petroleum Industry in Alaska

designed to produce virtually no residual oil to sell in today's

shrinking market.
Facilities using Cook Inlet natural gas. Three major
producing fields in the Cook Inlet area are the main support
of Southcentral Alaska's natural-gas industry. The gas from
these fields is "sweet" gas, gas that contains hardly any
hydrogen sulfide, carbon dioxide, or condensate (essentially
the same thing as NGL's --- hydrocarbons that are liquid
under atmospheric conditions), and production operations
are therefore relatively simple.
The North Cook Inlet field was discovered in 1962, but
the absence of a market delayed its development for several
years. Eventually, the Phillips Petroleum Company ar-
ranged to sell production from the field as liquefied natural
gas (LNG) to two Japanese utilities. In 1967, Phillips bought
out the other leaseholders and developed the field from a
single platform. The gas is piped to shore through two
undersea lines and then moves in a single line to the LNG
plant at Nikiski, where it is cooled and liquefied. The LNG
is then loaded into special "cryogenic" tankers, which ship
the equivalent of 11/0 mmcf/d to Japan.
The Beluga River gas field is not yet completely
developed. The gas in this field has been sold to an
Anchorage-based electric utility, the Chugach Electric As-
sociation (CEA), which uses it to fire combustion turbines at
Beluga on the west shore of Cook Inlet.
The Kenai field is a large gas field immediately south
of Kenai along the west shore of Cook Inlet. Most of the
field is onshore, on acreage owned by CIR! (Cook Inlet
Region, Inc., a Native corporation) as the result of a land-
swap with the State. Some of the gas from the Kenai unit is
produced for sale to the Alaska Pipeline Company (APC)
which carries it to the Anchorage gas utility, an affiliate of
APC. The balance is piped to Kenai, where it is used to
manufacture aqueous ammonia and urea fertilizer at a plant
operated by the Collier Carbon and Chemical Company (a
Union Oil Company subsidiary) on behalf of itself and Japan
Gas Chemical Company. Some of the Kenai gas is liquefied
The Petroleum Industry in Alaska Page 19

at the Phillips LNG plant, some is sold directly to the local

gas utility in Kenai, and the remainder is sent to the
Swanson River oil field where it is used to repressurize that
field to improve oil recovery.
The Pac-Alaska LNG project is a plan by two West
Coast utilities to liquefy Cook Inlet natural gas and ship it
as LNG to a terminal and regasification plant in California.
Actual construction of the project is now doubtful, because
of (!) the sponsors' inability thus far to get sales com-
mitments for the full volume of gas necessary to support the
plant, (2) a protracted contest before several regulatory
agencies --- now over, but succeeded by lawsuits --- over
the California terminal site, and (3) a growing abundance of
Lower-48 and Canadian gas that the California utilities can
obtain direct! y by pipeline.
Residental consumers, industry, and electric utilities
in the Anchorage-Cook Inlet region currently enjoy some of
the lowest natural-gas prices in the United States. The
average wellhead price in 1980 was about 27 cents per mcf
compared to a national average of $1.61 per mcf, and a
$4.91 border price for Canadian imports to the United
States. The provisions of the present sales contracts would
raise most Cook Inlet gas prices to whatever levels are paid
by the Pac-Alaska LNG plant, if that project is actually
built.
Prudhoe Bay Natural-Gas Reserves and ANGTS
The Alaska Department of Natural Resources (DNR) in
1980 estimated the proved natural-gas reserves of the
Prudhoe Bay Sadlerochit reservoir at 29 tcf, with 4.5 to 7 .8
tcf in other nearby reservoirs. Thus far, the natural gas
dissolved in the crude oil withdrawn from the Sadlerochit
reservoir is all being reinjected, except for a very small
quantity used as local fuel. After a natural-gas pipeline is
completed, the gas stream will then be stripped of water,
carbon dioxide, and most of its natural-gas liquids (NGL's),
and shipped through the pipeline to gas-transmission com-
panies in the Lower 48.
Page 20 The Petroleum Industry in Alaska

ANGTS --- The Alaska natural gas transportation

system. In 1977, the President and Congress awarded the
Akan Pipeline Company, a subsidiary of Northwest Energy
Company, the right to build the Alaska segment of ANGTS,
which will consist of a pipeline laid parallel to TAPS as far
as Fairbanks, whence it would follow the Alaska Highway
into Southern Alberta. There, the system branches into a
"Western Leg" to California, and an "Eastern Leg" into the
Midwestern States.
Alcan has now been succeeded by the Alaskan North-
west partnership; the Canadian sections would be built by a
group of companies operating under the name Foothills. The
Eastern Leg, known as the Northern Border system, is now
under construction by a partnership headed by InterNorth
(formerly Northern Natural Gas); the Western Leg is being
built by a subsidiary of Pacific Gas and Electric Company.
The ANGTS sponsors plan to design the system for an
inital thoughput of 2.0 bcf per day, beginning in 1986; they
have already received a number of important regulatory
approvals in both the United States and Canada, including
final authorization to "pre-build" the Southernmost sections
designed to carry Canadian as well as Alaska natural gas.
The 1977 presidential decision selecting ANGTS, however,
has several provisions that effectively block financing of the
rest of the more-than-$30 billion system. The pipeline
sponsors, the Prudhoe Bay gas producers, and their
respective financial advisors have asked Congress for
waivers to some of these provisions, but the fate of this
request is now (August, 1981) uncertain. Moreover, even the
proposed waivers would not solve all of the system's
organizational and financial difficulties, and it is unlikely
that the impasse will be resolved soon enough that the
pipeline can actually be built and completed on schedule.
The sales gas conditioning facility (SGCF). Natural
gas from the Sadlerochit reservoir is relatively "sweet" and
"wet" --- devoid of hydrogen sulfide, but saturated with
NGL's and has a high carbon-dioxide (C02) content
(about 13 percent). A "sales-gas conditioning facility"
The Petroleum Industry in Alaska Page 21

(SGCF) would reduce the level of C02 in the "sales gas"

(natural gas shipped through ANGTS) to a level consistent
with "pipeline quality" standards. Preliminary designs for
the SGCF, prepared for the gas producers and the ANGTS
sponsors, would use a physical (rather than chemical) pro-
cess called Selexol to remove C02 from the raw gas.
The Selexol process separates the components of the
produced-gas stream according to their different boiling
points. Two-thirds of the ethane that enters the condition-
ing plant stays in the sales gas to be shipped out in the gas
pipeline. Because the boiling point of ethane (C2H6) is close
to that of C02, however, some of the ethane remains mixed
with the C02 in a "waste gas", about half of which would be
used as plant fuel and the remainder returned to the field
for other local fuel uses. As the SGCF and nearby pumps,
compressors, and heaters must use some fuel or another in
large quantities, this arrangement may be an excellent one
if there is no better use for this portion of the ethane.
Prudhoe Bay ethane, however, may be most valuable
as raw material for an Alaska-based complex to produce
ethylene and its derivatives. This would be the case if the
ethane could be delivered to an appropriate plant site at an
acceptable cost after it has been extracted during or after
the conditioning process. The COz-removal process could
affect the amount of ethane available for petrochemical use
if the ethane is recovered downstream from the conditioning
plant instead of upstream. Even if the ethane were
extracted downstream, however, its volume would be more
than sufficient to supply a single "worldscale" ethylene
plant.
A natural gas liquids (NGL's) pipeline and Alaska
petrochemicals production. Prudhoe Bay natural gas con-
tains other NGL's in addition to ethane (C2): propane (C3),
butanes (C4), and pentanes-plus (C5+), each of which has
several alternative uses. Propane can be used directly as
home heating or industrial fuels in the form of "bottle-gas",
while propane and butane may be used along with ethane to
produce "olefins", such as ethylene, propylene, butylene, and
Page 22 The Petroleum Industry in Alaska

their derivatives. Butane may be used as the principal raw

material for methyl tertiary butyl ether (MTBE) and other
synthetic high-octane gasolines. The Exxon Chemical Com-
pany and the Dow-Shell group are independently studying
the economic feasibiJity of NGL's-based petrochemicals
production in Alaska.
The outlook for such a chemical industry is intimately
intertwined with decisions concerning Prudhoe Bay hydro-
carbons production and ANGTS. For example:
I. If ethane and heavier hydrocarbons are to be
recovered in sufficient quantities at Prudhoe Bay to
justify building an NGL's pipeline and ethylene plant,
the SGCF design may have to be modified.
2. If ethane and the heavier hydrocarbons are
recovered for use as chemical feedstocks, the energy
content of the recovered NGL's and the hydrocarbons
used for fuel in the ethane-extraction facility will not
be available for transportation through the natural-gas
pipeline and therefore might have an adverse impact
on the economics of ANGTS.
3. Exxon and ARCO, owners of the bulk of the
Prudhoe natural gas and NGL's, are major chemical
producers, and they must either be interested them-
selves in building (or participating in) an NGL's line
and an ethylene plant, or be willing to sell other
parties like Dow-Shell sufficient volumes of NGL's to
support both the NGL's pipeline and the petrochemical
facility.
t+. Whatever the resolution of these three issues,
the feasibility of building and operating an NGL's line
and an Alaska-based worldscale ethylene plant has yet
to be demonstrated; their operation raises some ad-
ditional questions. For example ---
5. A single worldscale ethylene plant would re-
quire only about 35 mb/d of ethane; some additional
ethane could conceivably be sold as electric utility
fuel in Interior and Southcentral Alaska, but the total
The Petroleum Industry in Alaska Page 23

assured demand within Alaska would be considerably

less than the volume of liquids (at least 150 mb/d)
necessary to justify construction of a new pipeline.
As shipment of surplus ethane beyond Alaska would
require cryogenic tankers similar to those used to move
LNG, large volumes of propane and butanes (which must also
be shipped under pressure or refrigeration, but require less
sophisticated tankers than LNG) would have to be saleable
in export markets in order to cover the pipeline cost, at
least until two or more ethylene plants were operating in
Alaska.
Finally, production of ethylene generates its own prob-
lems. Ethylene remains a gas unless it is chilled to -1550 F
(-1040 C); sometimes it is shipped by sea on a small scale in
cryogenic vessels similar to those used for LNG, but costs
probably rule out this strategy for a worldscale Alaska
facility. The chemical companies that have expressed
interest in producing ethylene from ethane in Alaska there-
fore contemplate processing the ethylene further into com-
pounds such as polyethylene, ethylene glycol, or stryrene,
which are solids or liquids under normal atmospheric con-
ditions and hence are easier to transport.
 lI
\
I

/ I I J, (___,- ,
,(J-V:
l

l __,.

'
I 1
 CHAPTER 3
FUNDAMENTALS OF HYDROCARBONS CHEMISTRY
General Introduction.
Fuels refining and petrochemicals manufacturing are
both hydrocarbons-processing industries. They begin with
mixtures of hydrocarbons from crude oil or natural gas as
raw materials, separate them into components, and alter the
molecules in various ways to produce a range of products for
final consumers or for use as inputs to other industries.
The refining and petrochemical industries overlap
technically, using many of the same processes and inter-
mediate products. The chief distinction between them is in
their respective "product slates". The greatest part of
refinery output is made up of liquid hydrocarbon mixtures,
along with certain solid byproducts of fuels refining, such as
asphalt or petroleum coke. While some refinery products
are sold for use as lubricants, solvents, or raw materials for
the petrochemical industry, the main business of the refin-
ing sector is fuels production.
Petrochemicals manufacturing includes practically any
hydrocarbons-processing operation whose principal output is
not liquid hydrocarbon fuels. Petrochemical products may
be liquid, gaseous, or solid: they include synthetic fibers,
plastic, paints and varnishes, resins, food additives, medi-
cines, industrial reagents, and much more.
Composition of Natural Hydrocarbons.
Natural hydrocarbons are complex mixtures of carbon
and hydrogen that are usually found underground in combi-
nation with impurities such as water, sulphur, and carbon
dioxide. Conventionally, hydrocarbons are grouped accord-
ing to the number (n) of carbon atoms (Cn) in each molecule.
However, the variations of hydrocarbon mixtures are vast
and every accumulation of oil and gas is unique.
Table I/ lists the names of some simpler, smaller-
molecule hydrocarbons found in crude-oil and natural-gas
reservoirs, and alludes to the existence of others with
dozens of carbon atoms in each molecule.
Page 26 Hydrocarbons Chemistry

Table 4
Elementary Hydrocarbon Compounds

Chemical Principal
Compound Formula Names
methane CH4 Natural gas.
ethane C2H6 Natural-gas
propane C3Hs liquids (NGL's)
butane C4HJQ or condensate.
pentane C5H12 Pentanes-plus,
hexane CGH14 natural gaso-
heptane C7H16 lines, or
octane CsH1s naphtha.

Oils, waxes, tars,

bitumen, asphalt.
etc.

The lightest and most stable hydrocarbon is methane

( CH4), the chief component of natural gas and a building
block for other hydrocarbons. Methane and ethane (C2H6)
are usually transported from the field in gaseous form and
sold to long-line gas transmission companies, which in turn
sell them to local gas distribution companies, most of
whose customers use gas directly as fuel without further
processing. The methane of natural gas is, however, also
frequently used as a feedstock to make synthesis gas for
processing into methanol, ammonia, urea, amines, and their
derivatives.
The bulk of the natural-gas liquids (NGLs), which may
or may not include the produced ethane, is usually separ-
ated in the field and sold for fuel use as LPG {liquefied
petroleum gas) or as feedstocks for petrochemical manu-
facturing.
The term "crude oil" usually refers to the heavier
hydrocarbon fractions, composed of molecules with five or
Hydrocarbons Chemistry Page 27

more carbon atoms. Crude oils are very complex mixtures

with many thousands of individual hydrocarbon compounds,
and range in consistency from natural gasolines to viscous,
semi-solid materials such as the bituminous tar sands of
Northern Alberta,
Each hydrocarbon compound in crude oil has its own
boiling temperature, with the heavier compounds (those
having a greater number of carbon atoms in each molecule)
boiling at higher temperatures and the lighter compounds
boiling at lower temperatures. Table 5 illustrates the
relationship between hydrocarbon boiling points and weights.

Table 5
Hydrocarbon Boiling Points and Weights

Boiling
Com12ound Formula Tem12erature Pounds/Gallon
Propane C3H8 -44 F 4.2
n-Butane C4H10 310 F 4.9
n-Decane C10H22 3450 F 6.1

Every crude oil contains a distinctive mixture of

hydrocarbon compounds, ranging from very light mixtures
with about 75 percent of the hydrocarbons in the gaso-
line/naphtha range (C5 to ClQ) to heavy oils that are solid
or nearly solid at atmospheric temperatures.
Crude oils also contain small amounts of sulfur, nitro-
gen, heavy metals, and other contaminants. The percentage
of sulfur varies from as low as 0.03 percent in some crude
oils from Bolivia and Argentina, to as high as 7 .3 percent in
oil from the Qayarah field in Iraq. Alaska's Cook Inlet
crudes, with 0.1 percent sulfur, are regarded as "sweet," or
very low-sulfur supplies. Other important sources of low-
sulfur crude oil are Alberta, Indonesia, Nigeria, and Libya.
Prudhoe Bay Sadlerochit crude oil, with about 1.0 percent
sulfur, is described as medium-sulphur or intermediate-
sweet. Quayarah crude oil is considered extremely "sour".
Page 28 Hydrocarbons Chemistry

(Strictly speaking, only crude oils containing free hydrogen

sulfide (H2S) should be called sour, but the term is common-
ly used to refer to the presence in significant quantities of
any compound containing sulfur.)
Chemistry.
The mixture of hydrocarbon compounds and the kind
and amount of impurities in a crude oil generally determine
its yield of gasoline, distillate fuels, lubricating oils, and
petrochemical feedstocks. To obtain these products, refin-
eries and petrochemical plants subject the hydrocarbon
mixtures to a number of processes that separate the com-
pounds into fractions or "cuts", remove the impurities,
recombine or convert the hydrocarbons into other forms,
and blend them into products for sale or further manufac-
turing. Equipment for altering the chemical structure of
hydrocarbons varies, however, among different refineries
and petrochemical plants, and will thus yield different
product slates even from the same crude oil.
The differing chemical composition of various natural
hydrocarbon supplies is reflected in different refining and
processing techniques, differences in product quality, and in
the manufacture of a wide range of different petroleum
fuels, chemicals, and synthetic products. The chemistry of
hydrocarbons is, therefore, an essential prelude to our
discussion of fuels refining and petrochemical manufactur-
ing.
Paraffins. Paraffins (also known as "alkanes") repre-
sent a large proportion of the hydrocarbons present in crude
oil. The paraffin series is composed of "normal" compounds
having straight chains of linked carbon atoms, and their
corresponding "isomers" (or "iso-alkanes") --- compounds
with the same numbers of carbon and hydrogen atoms, but
with branched-chain molecules. Both have the general
formula CnH2n+2i the names of individual hydrocarbons in
the series end with "-ane". Methane and ethane are the
simplest paraffins, having the following structures:
Hydrocarbons Chemistry Page 29

H H H
I I

H-C-H H-C-C-H

H H H

Methane (CH4)

Similar! y, propane is:

HHH
I I I

H-C-C-C-H
I I I

H H H

Propane (C3Hg)

Hydrocarbons containing more than three atoms of

carbon in each molecule may form "isomeric," branched-
chain forms. Contrast, for example, "normal" (n-) butane
and "iso- 11 {i-) butane:

H
H H H H H-C-H
I I

H-C-C-C-C-H H-C-C-C-H
I I II II II

H H H H

Normal or Iso- or
n-butane i-butane
(C4H10) (also C4H 10)

Butane has only these two isomers. As the number of

carbon atoms increases, however, the number of possible
Page 30 Hydrocarbons Chemistry

structural combinations increases geometrically. For in-

stance, pentane (C5HJ2) has three isomers, nonane (C9H20)
has 35, and dodecane (C12H26) has 355.
Although a paraffin and its isomers have the same
number of atoms, they boil at different temperatures, have
different specific gravities, and participate in different
chemical reactions.
Naphthenes. Naphthenes are hydrocarbons with more
than four carbon atoms arranged in ring-like central struc-
tures rather than straight or branched chains, and have the
general formula CnH2w (For simplicity, we have omitted
the H symbols in the following schematic diagrams. )

c
c c c
c c
c c
C-C c
Cyclopentane Cyclohexane
(C5H10) (C6H12)

At least one five- or six-membered ring is present in

every naphthene. Cyclopentane and cyclohexane are the
only hydrocarbons in the series that occur in nature. The
more complex members of the series consist of one or more
central rings, with one or more paraffin-like branches
("alkyl" groups) attached to them. The number
of compounds which, in the course of refining processes,
may attach in different combinations to the outside of the
ring can be very large, however. One such compound is
methyl cyclopentane:
Hydrocarbons Chemistry Page 31

c
C C-CH3
C-C

Methyl
Cyclopentane
(C6H12)

Aromatics. The simplest member of the aromatic

series and the building block for all other aromatics is
benzene, composed of a six carbon-atom ring like cyclo-
hexane, but with only six associated hydrogen atoms, and
with three single and three double bonds alternating bet-
ween the carbon atoms in the ring:

c
c c
II

c c
c

The aromatics include all compounds whose molecules

contain at least one benzene ring. Some such compounds
are formed by substituting paraffin units for one or more of
the hydrogen atoms in the benzene molecule. These hydro-
carbons are called alkyl benzenes; one example is toluene.
Other compounds, like naphthalene, contain more than one
benzene ring:
Page 32 Hydrocarbons Chemistry

CH3
I

c c c
c c c c
II
c
II

c c c c c
c c c
Toluene Naphthalene
(C6H5CH3) (C10Hg)

The substitution of double carbon bonds for hydrogen

bonds is easy to understand if one remembers that each
carbon atom in a larger molecule almost always has four
links or bonds with other atoms. If hydrogen atoms are
removed from a hydrocarbon molecule, the carbon bonds
that are left empty tend to link with one another to create
double bonds. The double bonds in the benzene ring are very
unstable and chemically reactive, however, and thus the
members of the alkyl benzene series are important building
blocks for refined petroleum products and petrochemicals.
Olefins. The olefins are not found in crude oil, but
are manufactured from oil, natural gas, or NGL's by one of
several cracking processes. They resemble paraffins and
naphthenes in structure, but like aromatics they have dou-
ble (and sometimes triple) bonds between carbon atoms.
The double and triple bonds are deceiving because,
contrary to appearances, these bonds are weaker than a
single bond, making the compound unstable. If every car-
bon bond were linked to an atom of hydrogen (or some other
element), the hydrocarbon would be "saturated" and there-
fore relatively stable. Olefins and aromatics are said to be
"unsaturated" because they contain double or triple bonds.
Hydrocarbons Chemistry Page 33

H
I

H H H-C-H
H-C C-H
C=C
I I
H H HC=CH
I I

HH

Cyclopentene
(C5Hg)

The unsaturated hydrocarbons are valuable to the

chemical industry precisely because they readily react di-
rectly with other chemicals to form more complex com-
pounds. For instance, the olefin ethylene (CzH4) reacts
with chlorine to form vinyl chloride monomer, (VCM),
which in turn is used to produce polyvinylchloride (PVC)
resin used for the manufacture of plastics.
/ I

,' i
/

I ,/

v
 CHAPTER 4
FUELS REFINING
The main business of the refining sector is fuels
production. The manufacture of refined fuels begins with
natural hydrocarbon compounds and separates them by dis-
tillation, tears them down, and rebuilds and restructures
their molecules into produce saleable products. Before the
l 900's, a typical refinery just distilled the crude oil into a
series of "cuts" or fractions, which were sold as straight--
run fuels. Today, almost all petroleum products are speci-
ally tailored in their physical and chemical properties and
freedom from impurities to meet exacting market demands,
thus requiring treatment that extends far beyond simple
distillation.
Petroleum Industry Structure.
The refining sector is an integral part of a petroleum
industry made up of thousands of companies that are ex-
ceedingly varied in size, functions, geographical sphere of
operations, and structure.
The major oil companies. "Big Oil" consists of seven,
twelve, sixteen, or twenty "major" or "multinational" corpo-
rations, depending upon the statistical authority. However
many "Sisters" one chooses to count, the major oil compa-
nies are distinguished by both their great size and their
vertical integration: They produce crude oil; own crude-oil
and petroleum-product pipelines, tankers, and barges; refin-
eries; tank farms and terminals; and operate retail outlets.
Many of the majors are engaged in other related businesses,
such as natural-gas production and processing, and petro-
chemicals manufacturing. These major companies vary
greatly in size, and no two of them have the same mix of
functions; some majors are net sellers and others net buyers
of crude oil; some are net sellers of refined products at
wholesale and others net buyers, and in many different
degrees.
In 1979, the top 16 integrated companies produced
about 60 percent of U.S. crude-oil output and accounted for
about 12 million barrels per day (mmb/d) of refining capa-
city, or about two-thirds of the national total. The same
Page 36 Fuels Refining

companies also marketed about two-thirds of the refined

products sold in the United States.
The independents. A significant part of the business in
each sector of the petroleum industry is conducted, how-
ever, by "independents" --- specialized or only partially-
integrated firms that compete both with the majors and
with one another. There are independent exploration com-
panies and producers; independent oilfield service companies
and gathering companies, independent oil-pipeline and tan-
ker-transportation companies, independent refiners, resel-
lers, and brokers; jobbers, marketers, and retailers.
The independent sector is deeply rooted in U.S. oil-
industry history. From its earliest days, the production of
crude oil in the United States was widely dispersed among
many producing companies, largely because oil was dis-
covered in fields of many sizes located on privately-owned
tracts where farmers, ranchers, and other owners held the
subsoil mineral rights as well as the surface estate. Al-
though the top 20 integrated oil companies have acquired
control of about two-thirds of the crude-oil output in the
United States and three-fourths of the reserves, many fields
have several operators and royalty owners. Data from
Windfall Profits Tax collections reveal that the United
States has literally tens of thousands of crude-oil producers
and about two million royalty owners.
The majority of oil-field discoveries onshore in the
Lower-48 appear to have been made by independent "wild-
cat" exploration companies, and they continue to contribute
a smaller yet significant portion (about one-third) of the
new crude-oil reserves added annually. Because their cost
structures and exploration strategies differ from those of
the majors, there is a tendency for independent exploration-
ists to sell their discoveries to major producers, while the
majors often sell off nearly depleted fields and hlgh-cost
"stripper-well" (wells producing less than JO barrels per day
production) properties to specialized independents.
The situation is somewhat different on the Outer
Continental Shelf (OCS) and Alaska. There, the ownership
Fuels Refining Page 37

of prospective petroleum acreage is concentrated in the

Federal and State governments, and lease tracts are much
larger than the typical southwestern U.S. farm property. In
these areas, the high costs of exploration tend to restrict
activity to the major companies and joint ventures of the
larger independents. Even so, OCS and Alaska State lease
auctions typically attract 10 to 50 different bidding combi-
nations, representing dozens of separate companies.
About 6 mmb/d or 31/ percent of the total U.S.
refining capacity were owned by non-integrated refining
companies in 1979. As one might expect, the independent
refiners depend far more heavily on crude oil from indepen-
dent producers than do the refining divisions of the major
companies. In retailing, the majors tend to sell their own
refined products, or refined products exchanged with other
majors, under their respective brands, while independent
marketers buy their products at wholesale from major
companies, independent refiners, and resellers.
Feedstocks and Petroleum Products
Within rather narrow limits, the characteristics of a
refinery's crude-oil supply and its initial design determine
the possible mix of its refined product output. Refineries
are planned, therefore, to match their product slates as
closely as possible to the mix of product demand in the
areas the refinery serves. North American refineries, for
example, have been generally designed to emphasize gaso-
line production, and secondarily, that of "middle distillates"
(heating oil, diesel fuel, and jet fuel), at the expense of
heavy fuel oils.
Closer to home, Chevron's Kenai refinery processes
crude oil to serve local markets for jet fuel, diesel fuel, and
home heating oil. Mapco's North Pole refinery near Fair-
banks cuts the "tops and bottoms" (the lightest and heaviest
fractions) out of the crude oil, in order to sell the middle
distillates, and the Tesoro refinery produces gasoline and
middle distillates. Each of them, however, exports a large
part of each barrel to other states in the form of residual
oil, for which there is no significant demand in Alaska. Had
Page 38 Fuels Refining

it been built, Charter's proposed Alaska Oil Company re-

finery at Valdez would have been the state's first "complex"
refinery, capable of processing all the residual oil from the
distillation tower into lighter refined products.
Refinery design also reflects the grade and quality of
crude oil to be processed. Refinery complexity, fixed costs,
and operating costs depend principally upon the match or
mismatch between feedstock characteristics and the pro-
ducts to be produced. Thus, light (high-gasoline) and sweet
(low-sulfur) crude oils have long been preferred refinery
feedstocks in North America, where motor fuels have been a
relatively large part of total petroleum demand and where
air quality has been a major concern. Fortunately, the
grade and quality of North American crude oils (other than
in California) have tended to be well matched to domestic
product slates.
Feedstock characteristics. The characteristics of dif-
ferent crude oils determine, to a large extent, the refinery
processes needed to make a particular product slate. Each
crude oil is unique, yielding different amounts of and
different mixtures of compounds within each fraction.
These characteristics are ascertained by means of a crude-
oil assay involving controlled fractionation in the laboratory
and the chemical analysis of each fraction. The assay
results typically describe a crude oil in terms of the
proportion of its total weight falling into each straight-run
fraction, and its density, sulfur content, viscosity, pour
point, metal content, and often the proportion of straight-
line paraffins, branched-chain paraffins, naphthenes, and
aromatics.
Density is the ratio between the weight of a substance
and its volume, for example, kilograms per liter or pounds
per barrel. For crude oil, density serves as an index of the
relative proportions of the different hydrocarbon fractions,
with the compounds that contain the largest number of
carbon atoms per molecule having the greatest density, and
the smaller-molecule LPG's and natural gasolines the least.
The density measure is also affected by the proportions of
Fuels Refining Page 39

the four major hydrocarbon types, as the individual densities

of compounds with a given number of carbon atoms per
molecule is greatest for the aromatics. Naphthenes, iso-
paraffins, and normal paraffins have progressively lower
densities
A low-density crude oil can yield more than half of its
weight in light distillates (straight-run LPG's, gasoline,
kerosene, and naphtha), while there are high-density Cali-
fornia crudes in which these lighter fractions comprise as
little as 6 percent of total weight. Alaska North Slope
(Sadlerochit) crude oil is somewhere in the middle with
about 30 percent light distillates. Density can be measured
in terms of specific gravity (the ratio of the weight of a
given volume of a substance to that of an equal volume of
water), but the petroleum industry generally prefers to use
"AP! gravity", a measure denominated in degrees in which
lighter or low-density crude oil is referred to as having a
"high AP! gravity", in a confusing violation of the layman's
common intuition. A high-density crude oil is similarly
referred to as having a "low AP! gravity." Some heavy
California crudes have AP! gravities in the JOO to 160 range;
Prudhoe Bay crude oil has an AP! gravity of 270; and "light"
crude oils from Cook Inlet have AP! gravities as high as ti!O.
The total sulfur content of a crude oil is measured in
terms of its proportion of sulfur(s) in the total weight of the
crude oil, and thus the volume of sulfur compounds likely to
be present in the products refined. Because sulfur atoms
have an affinity for the heavier hydrocarbon molecules, the
heavier crude oils generally (but not always) tend to have a
higher sulfur content. Cook Inlet, Alberta, and Nigerian
crude oils tend to be have a relatively low sulfur content of
less than 0.3 percent; Prudhoe Bay crude oil is regarded as a
medium-sulfur product at about I percent, while some
"sour" California crudes contain more than 3 percent sulfur.
Since 80 to 90 percent of the sulfur typically remains
in the "residuum", (the substance that remains in the bottom
of the refinery fractionally lower after the lighter products
have boiled off), the acceptability of heavy fuel oils under
Page 40 Fuels Refining

prevailing air-quality standards depends chiefly on the sulfur

content of the crude oil. High-sulfur crudes tend to leave
impermissable amounts of corrosive and polluting sulfur
compounds in the middle distillate as well, requiring costly
hydrotreating before the products can be marketed. Proces-
sing high-sulfur crude oils also requires special catalysts and
more sophisticated refinery metallurgy, meaning that a high
sulfur content in the refinery feedstock makes it consider-
ably more costly to convert into a given slate of refined
products.
Generally speaking, crude-oil types and qualities are
categorized as follows:

Atmospheric Residuum by Weight@ 10500 F

Sulfur by Weight Less then 15% More than 15%
less than 0.5% light low-sulfur heavy low-sulfur
0.5% to 1.0% light medium-sulfur heavy medium-sulfur
more than 1.0% light high-sulfur heavy high-sulfur

Viscosity, pour point, and wax content indicate

how easily crude oil will flow through pipelines and into or
out of tanks and tankers, and the degree to which solid
deposits are likely to build up on pipeline or storage-tank
walls. All of them are, therefore, crucial variables in
designing pipelines and storage facilities. Pour point is the
lowest temperature at which oil will pour or flow in
response to gravity, under standard conditions. Examples of
pour points are:

Bonny Light (Nigeria) +50 F

Prudhoe Bay Sadlerochit _50
Saudi Arabian Light -30

Viscosity is a measure of the resistance to flow in a

liquid at a given temperature and pressure, and increases as
Fuels Refining Page 4-1

the temperature increases. A high wax-content crude oil

like Indonesian Minas crude tends to clog pipelines, so that
they have to be "pigged" (scraped out by a special
device sent through the line) frequently.
In addition to these characteristics, there are other
features of crude oil that affect its product yield and cost
of refining. The most important of these characteristics
are the relative proportions of paraffinic and naphthenic
hydrocarbons, and the metals content.
Refinery Products.
Refined products include a full spectrum of intermed-
iate and consumer products:
First-stage products. Distilling crude oil into frac-
tions is the first step in all petroleum-refining operations.
This process yields a set of straight-run "cuts" or product
mixes that are the intermediate building blocks for refined
products. These fractions are characterized by their boiling
ranges --- the hydrocarbons with the lowest boiling points
being the lightest compounds.
Table 6 illustrates the relationships among cut points,
straight-run fractions, and refinery end-products. Each of
the various end products is composed of hydrocarbons having
a rather broad range of boiling points, while different end
products have boiling ranges that overlap. As a result,
refiners are able to vary the proportions of different pro-
ducts made at a given refinery by varying the temperatures
or cut-points. For example, adjusting refinery operations to
raise the cut-point temperature at which straight-run gaso-
lines are separated from naphtha means that (I) less gaso-
line and more naphtha will be produced (perhaps for use as
military jet fuel), and (2) the produced gasoline and naphtha
will both be lighter than before.
Distillation of two different crude-oil types in identi-
cal refinery facilities will, moreover, yield gasoline of dif-
ferent octane ratings and a light gas-oil fraction of differ-
ent cetane ratings (a measure of diesel-fuel quality com-
parable to octane ratings for gasoline). Thus, the amount of
Page 42 Fuels Refining

Table 6
Crude-Oil Distillation

Hydrocar- Temperature Distillation

bon type cut-Eoints Eroducts End-Eroducts
C1 methane Natural gas
C2 less than
C3 1000 F LPG LPG
C4 100 Motor gasoline
C5 150 straight- Naphtha petro-
200 run chemical feedstock
through gasoline
250 Military jet fuel
CJO 300 naphtha Civilian jet fuel
C10+ 350 kerosene
No. 1 diesel &
400 stove oil
450 No. 2 diesel &
500 light stove oil
550 gas oil No.4
600 turbine fuel
650
700 heavy Gas-oil
gas oil petrochemical
750 feedstock
800
850 Residual fuel
residuum oil
900
950
Bunker "C"
1000
more than
10000 F Asphalt
coke
Petroleum coke
Fuels Refining Page 1/3

reforming and other processing required to turn different

crudes into marketable products varies widely.
End products. Refinery end-products can be grouped-
as follows:
Motor gasoline. At one time, light naphtha
fractions direct from the distillation tower were sold as
straight-run gasoline; however, today's cars would run very
poorly on such fuels. Refiners have altered the composition
of gasoline considerably by means of reforming, blending,
and compounding with additives, in order to control pre-
mature ignition and detonation ("knocking"), vapor pressure,
gum formation in the engine, odor, and overall performance.
For several decades, refiners have produced and mar-
keted at least two octane levels of leaded gasoline (regular
and premium). Since the early l 970's, changes in automobile
design intended to reduce air pollution have forced refiners
to offer, in addition, at least one grade of unleaded gasoline;
the sale of premium leaded gasoline is now being phased out
with the decline in the number of cars that require it.
These changes in gasoline requirements result in more
complex and expensive refining operations and reduce the
amount of gasoline that can be obtained from a barrel of oil.
Diesel fuel. Refineries manufacture diesel fuel
for high-speed stationary, highway and marine diesel engines
from the middle-distillate fractions of the crude oil. Fuel-
quality requirements depend largely on engine rotational
speeds. Fuel for high-speed diesel engines is made from the
lighter portions of the distillate cut, and overlaps to some
extent with kerosene.
Engines used for electrical generation or marine pro-
pulsion run at lower rotational speeds than automotive
engines and will accept a lower quality fuel. A marine
diesel fuel, therefore, often consists of a blend of distillates
and heavy gas oil.
Like motor gasoline, distillate diesel fuels for use in
automotive engines have improved during the past several
years to meet requirements imposed by changes in engine
design and operation. The most significant change in diesel
Page 41/ Fuels Refining

fuels has been the use of hydrogen treating in refineries,

primarily to reduce sulfur content. Fuels have also been
improved to decrease engine deposits and reduce smoke and
odor. The use of additives in diesel fuels has become
common for the purpose of lowering "pour points" (insuring
that the fuel continues to flow at low temperatures),
increasing stability in storage, and improving the ease of
ignition.
Aviation fuel. Aircraft fuels are of two quite
different kinds: aviation gasoline ("Avgas") for piston--
engined craft, and jet fuels for use in turbine engines.
Aviation gasoline generally requires higher antiknock ratings
than motor gasoline and, because of the greater range of
atmospheric pressures and temperatures, more exacting
vapor-pressure standards.
A satisfactory turbine fuel must ignite easily and burn
clean! y; and because jet fuels are exposed to very high and
low temperatures in use, they must therefore have low
freezing points and at the same time be stable at high
temperatures. These qualities are less demanding on refin-
ery design and operation, however, than those that are
critical in fuels for internal-combustion engines. As a
result, marketable jet fuels can be produced even in rela-
tively simple refineries, like Mapco's North Pole plant, and
tend to be cheaper to manufacture than the same amount of
energy in the form of Avgas.
An alternative jet fuel used mainly by the military is
known as "wide-cut" gasoline and is, as its name suggests, a
product blended from straight-run fractions ranging from
the light naphthas to heavy gas oil (but mainly the former).
This fuel, known as "aviation turbine gasoline" or JP-4, is
easily manufactured, and because of its wide cut, refiners
can obtain a high yield from each barrel of crude oil.
Gas and LPG (liquefied petroleum gas). Various refin-
ing processes liberate considerable volumes of gaseous
hydrocarbons (methane, ethane, propane and butanes).
These gases are typically used as fuel within the refinery
itself. Refinery gases, particularly methane and ethane, are
also important feedstocks for the manufacture of petro-
Fuels Refining Page 45

chemicals, including methanol, ammonia, ethylene and their

derivatives. Butane and isobutane are blended directly into
motor gasoline to increase its vapor pressure and, hence, to
assure that it will ignite.
The butanes and propane ("liquefied petroleum gases"
or LPG) released during refining also become feedstocks for
certain intermediate processes in the manufacture of motor
gasoline and additives like MTBE (methyl tertiary butyl
ether), which raise the octane rating of gasoline. Under
moderate pressure propane remains liquid at ambient tem-
peratures, and can therefore be marketed safely as "bottle
gas" for space heating and cooking. Gas utilities also mix
propane and butane with air to form an additive or substi-
tute for natural gas during peak-demand periods, and there
are a large number of industrial uses for propane, including
metal cutting and welding using oxy-propane torches, and as
process fuels.
Distillate fuel oil. Distillate fuel oil includes the Nos.
1, 2, and 4 heating oils; and the term is often used to include
diesel fuels as well, which are almost identical to distillate
heating oils. No. 1 stove oil is the lightest of the distillates
and, because it remains liquid and ignites readily at very low
temperatures, is the main home-heating fuel in Alaska's
interior. No. 2 heating oil is the most common home and
commercial heating oil nationally and worldwide. The price
of No. 2 fuel oil is a frequently used indicator of petroleum-
product costs.
Since World War II, refiners have improved the quality
of distillate heating oils by removing sulfur and nitrogen
through hydrogen treating and reducing the quantity of ash
or other deposits left when the fuel is burned. Just as they
do for gasoline and diesel fuels, refiners ad just the hydro-
carbon blend in each grade of distillate heating oil to match
the particular season and location.
Residual fuels. Residual fuels are made from the
heaviest hydrocarbon fractions and are commonly marketed
as Nos. 5 and 6 heating oils, heavy diesel, heavy industrial,
and Bunker C fuel oils. Residual fuel oil has a higher energy
content per unit of volume (e.g., per gallon) than other
Page 46 Fuels Refining

petroleum fuels, but it must be heated before it will flow

through a pipe or burn in a furnace or turbine. Typically,
therefore, these fuels are used to provide steam and heat
for industry and large buildings, to generate electricity, and
to power marine engines.
Residual fuel oil is therefore competitive with natural
gas and coal as in industrial and electric utility fuel
markets. While there are serious regulatory obstacles to
using either gas or coal as an electric utility and industrial
boiler fuel, the rapid runup in crude-oil prices since 1973 has
tended to make residual oil more valuable as intermediate
products for the manufacture of gasoline and distillate fuel
oils. Relative prices therefore increasingly favor (!) substi-
tution of coal, natural gas, and nuclear energy for residual
oil as industrial fuel and, (2) petroleum-industry investment
in new crackers and cokers to break up the residuum into
lighter hydrocarbon mixtures that can be processed and sold
for higher pr ices.
Lubricants. Lubricants are a diverse group of special-
ly-blended products falling into three general categories:
automotive oils, industrial oils, and greases. Engine oils,
gear oil, and automatic transmission fluids are three major
lubrication products used in automotive operations. These
products function to lubricate, seal, cool, clean, protect,
and cushion metal parts. Industrial oils are blended to
perform a variety of functions, including lubrication, fric-
tion modification, heat transfer, dispersancy, and rust-
prevention. Greases are basically gels and are composed of
lubricating oil in a semi-rigid network of gelling agents such
as soaps or clays.
Petroleum solvents. Although they represent a much
smaller market than, say, motor fuels, petroleum solvents
are made in many grades for a variety of uses. Solvents are
a major component of paint thinner, printing inks, polishes,
adhesives and insecticides, and are used for dry cleaning.
The manufacture of these products requires careful refining
to remove unwanted odors and maintain consistent product
quality.
Fuels Refining Page 47

Asphalt. The heaviest fractions of many crude oils

include natural bitumens or asphaltenes and are generally
called asphalt. Long before its use as a fuel, petroleum was
valued for its asphalt, which has been used throughout
recorded history. Because of its adhesive, plastic nature
and waterproofing qualities, it is widely used for road-
making.
Product mix. Individual refineries have considerable
discretion in the product slates they produce, even from a
single mix of crude-oil feedstocks. For this reason, it is
important to understand the factors that influence product-
slate decisions; these factors include -- in no particular
order of logic ---
* Feedstock assay and straight-run fraction mix.
* Crude-oil supply conditions.
* Refined product market conditions.
* Refinery flexibility regarding product slate.
* Refinery flexibility regarding feedstock mix.
* Refinery size and affiliation.
* Finished-product specifications.
Feedstock assay and straight-run fraction mix. The
discussion of first-stage products has already shown that the
hydrocarbon composition of crude oil determines the vol-
umes of different straight-run fractions into which the
crude oil can be separated by simple distillation.
Crude-oil supply conditions also influence refinery
design and product slates. For example, Mapco's refinery at
North Pole runs medium-gravity (AP! 270) Prudhoe Bay
crude oil into a simple atmospheric distillation facility. The
straight-run gasoline distilled from this crude oil is not
suitable for automotive use, but the straight-run naphtha
and light gas oil can be blended to make marketable jet fuel,
diesel oil, and home-heating oil. The lightest products, the
heavy gas oil, and the residual oil are mixed with the
remaining Prudhoe Bay crude oil flowing through (TAPS) and
delivered to more complex refineries in the Lower 1;8 that
can process these hydrocarbons into marketable products.
Page lf8 Fuels Refining

Elsewhere, a refinery like the North Pole plant would

probably have a hydrocracker to convert most of the heavy
gas oils coming from the fractionating tower into gasoline
or middle distillates. But in Fairbanks, the processing and
marketing of the heavy gas-oil fraction would leave the
remaining residual oil with an AP! gravity lower than the
]70 minimum established by the Alyeska Pipeline Service
Company for shipments through TAPS. Thus, if Mapco
wanted to convert more of its Prudhoe Bay feedstocks into
products that are saleable in Alaska, it must either (I)
process all of the gas-oil and residual-oil fractions into
lighter products at a very high cost, or (2) develop currently
nonexistent local markets for naphtha and heavy fuel oils.
Refined-product market conditions play a crucial role
in determining product slates. A given volume of petroleum
is generally cheaper to ship Jong distances in the form of
crude oil than as a diverse and varying assortment of refined
products. For this reason, transportation economics normal-
ly lead refineries to locate near their product markets and
each refinery, in turn, is normally designed to produce a
product slate that corresponds to local demand.
The mix of petroleum-product demand tends to vary
geographically according to a region's climate, level of
economic development, industrial character, and supply of
competing fuels. U.S. West Coast refineries have been
designed largely to produce transportation fuels such as
motor gasoline and jet fuels, because of (I) the region's mild
climate, (2) the mobility of its population, and (3) relatively
abundant regional supplies of natural gas and hydroelectric
energy. In the Northeast, on the other hand, climate,
lifestyles, and energy costs combine to encourage relatively
greater dependence upon heavy fuel oils. The design of
refineries in the two regions reflects these differences in
demand mix.
Product demand also varies seasonally. Gasoline con-
sumption typically peaks in the summer, but winter is the
peak season for home heating oil. Refineries are usually
designed with only enough flexibility to accomodate a part
of this seasonal swing in demand, because increasing the
Fuels Refining Page 49

product-slate flexibility beyond a certain level comes only

at increasing costs. For this reason, the seasonal supply
strategy of major refiners also involves "winterfill" and
"summerfill" --- putting the products into storage for sale
when the demand pattern reverses itself.
Different types of fuels require differing degrees of
precision in their product specifications. The performance
of industrial and electric-utility boiler fuels, for example, is
relatively insensitive to the exact character or size of
hydrocarbon molecules burned. Product specifications for
middle distillates --- stove oil, diesel fuel, and jet fuels ---
focus on easy ignition, clean burning, pour points and vapor
pressures. The demands these specifications make on re-
finery design and operation are rather moderate, because
there is a broad range of straight-run hydrocarbon blends
that are able to meet the requirements for any of these
fuels. Motor gasolines, however, have to be more closely
controlled with respect to molecular structure and im-
purities in order to assure ignition and to avoid vapor lock,
knocking, and unacceptable engine wear.
Aviation gasolines must meet the most severe product
specifications of any petroleum fuel, both because of the
extreme combustion conditions encountered in high-per-
formance piston engines and the potentially disastrous con-
sequences of engine failure. It is probably the risk of legal
liability from alleged quality shortcomings that has so far
deterred any Alaska refiner from producing Avgas for local
consumption, despite the relatively high demand for the
product in the state.
Refinery flexibility. Adding a hydrocracking or coking
unit to an existing refinery enhances its processing flexibil-
ity by permitting the upgrading of straight-run residuum and
heavy gas oils into gasoline and middle distillates.
Tesoro recently installed the first hydrocracker in
Alaska at its Kenai plant. The refinery was originally
designed to run light Cook Inlet crude, but as the supply of
that feedstock declined, Tesoro was faced with the choice
of (I) cutting back production accordingly, (2) running the
heavier .Prudhoe Bay crude oil, and thus producing less
Page 50 Fuels Refining

gasoline and middle distillates and more residual oil to be

exported from Alaska because of the lack of a local market,
or (3) adding equipment to upgrade the greater quantities of
residual oil produced by distilling Prudhoe Bay crude.
Tesoro chose the third alternative, installing a hydro-
cracker to process about 7 ,500 bpd of heavy gas oil ---
about 11 percent of the crude oil input to the refinery ---
into motor gasolines, jet fuel, and diesel fuel. Plunging
demand for residual coupled with a fall in the average AP!
gravity of crude-oil inputs is encouraging refiners to take
similar action everywhere in the United States. The Oil and
Gas Journal reported an increase in total U.S. hydrocrack-
ing capacity of close to 30 percent between year-end 1979
and year-end 1980.
Refinery size and affiliation. Independent refineries
in the United States with less than 30 mb/d capacity ---
especially the "bias-babies" spawned by the federal entitle-
ments system between 1973 and 1980 --- are typically
simple atmospheric distillation units producing a relatively
large proportion of residual oil and heavy refined products.
Not only do larger refineries tend to be more complex and
more flexible with respect to both feedstocks and product
slates but, all other things being equal, a large company
with many refineries has greater system-wide flexibility
because of its ability to produce different product slates in
different plants equipped to complement one another.
Of all the refineries operating in Alaska, for example,
the Chevron Kenai facility has from the beginning taken
advantage of California standard's system-wide flexibility.
Much of the heavy gas oil from Kenai is sent, along with the
residual oil, to the <company's Richmond plant, which already
processes Prudhoe Bay crude oil that Chevron buys from
Sohio. In the face of surplus system-wide capacity, more-
over, Chevron recently suspended production of military jet
fuel in Alaska, instead choosing to ship the straight-run
gasoline from Kenai to its El Segundo refinery for conver-
sion to benzene.
Fuels Refining Page 51

Refining of Petroleum
Distillation. All refinery operations begin with the
distillation of a crude-oil feedstock into petroleum frac-
tions. The crude oil can either be heated through a series of
temperature steps and the vapors condensed at each step, or
a large portion of the crude oil can be vaporized and the
vapor cooled in a series of temperature steps. Either way,
the crude oil is separated into fractions, each composed of
hydrocarbons having similar boiling-points. The boiling
point ranges of the more common products are shown in
Table 7 below.
Table 7
Boiling Ranges and Distillation Products

Boiling Rangeo F Product

less than
90 propane/butane
90-220 gasoline
220-315 naphtha
315-450 kerosene
450-800 gas oil
more than
800 residuum

In a typical refinery, the crude oil is heated to about

6500F as it enters the atmospheric distillation tower. The
vapors rise in the tower, are cooled and condensed on trays
at various levels and then withdrawn. Those heavy portions
that do not vaporize are withdrawn at the base of the tower
and sent to a vacuum distillation tower. Under reduced
pressure, additional hydrocarbons vaporize, rise in the se-
cond tower, and are separated as the vapors cool. The
heavy residue remaining is withdrawn at the base of the
vacuum tower.
Restructuring hydrocarbon molecules. The separated
fractions undergo further processing. Typically, the "light
ends" from the top of the fractionating column go to the gas
plant for further fractionation; the straight-run gasoline is
Page 52 Fuels Refining

blended into motor gasolines and jet fuel; naphtha is sent to

the reformer for processing, kerosene to a hydrotreater for
clean-up, light gas oil to distillate-fuel blending, heavy gas
oil to the cat cracker; and straight-run residue is fed to the
flasher.
Beyond distillation, refiners restructure or "reform"
the hydro-carbon molecules either by making the molecules
smaller or larger or by rearranging the molecular structure
of a hydrocarbon without changing the number of atoms. In
restructuring molecules, extensive use is made of catalysts,
substances that cause an acceleration of a chemical reac-
tion without themselves being permanently affected. Some
catalysts offer a surface structure that increases the rate of
reaction, and others may cause certain reactions that
would not otherwise occur. In many refining processes, the
use of different catalysts results in a different yield, such
as a different proportion of paraffins and aromatics. As a
consequence, the refining and petrochemical industries are
continually searching for new and superior catalyst mater-
ials.
Various processes have been given different names by
their inventors, but basic refinery operations can be classi-
fied into the categories shown in Table 8:
Cracking. When hydrocarbons are heated to tempera-
tures exceeding about 4500 C (81/20 F), some molecules
begin to break down or split. The reaction is very complex
and a number of different products are formed, including
heavier products as well as the predominantly lighter pro-
ducts.
In cracking, refiners heat a mixture of heavy hydro-
carbons to a high temperature under pressure. This process
causes the larger molecules to split; the result is a new mix
of molecules, but one with a much higher proportion of
lighter hydrocarbons, from methane through the gasoline,
naphtha, and middle-distillate ranges.
Fuels Refining Page 53

Table 8
Refinery Processes to Restructure Hydrocarbons

Process Basic Function

CRACKING: Breaking {"cracking") large mole-
cules into smaller ones. (Cracking
also produces some larger mole-
cules.)
REFORMING: Dehydrogenation --- removal of
hydrogen --- for example, convert-
ing saturated straight-chain hydro-
carbons into unsaturated aroma-
tics.
POLYMERIZATION Combining smaller molecules into
and ALKYLATION: larger ones; polymerization com-
bines identical molecules while al-
kylation combines different-type
molecules.
HYDROGENATION or
HYDROTREATING: The addition of hydrogen to con-
vert unsaturated to saturated
hydrocarbons, or to replace various
chemical radicals with hydrogen.
ISOMERIZATION: Rearrangement of the structure
within a molecule without changing
the number of atoms.
TREATING: Converting a contaminant into an
easily removable or non-objection-
able form.
COKING: A form of thermal cracking con-
ducted under high pressure, promo-
ting the formation of coke as well
as lighter products.
Page 51/ Fuels Refining

As large molecules break up through cracking, the lack

of sufficient hydrogen atoms to saturate all the carbon
bonds forces some of the carbon atoms to bond to one
another, forming olefins, smaller aromatic and naphthenic
rings, and coke. The lighter products of this process are
important chemical feedstocks --- ethylene, propylene and
butylenes. However, the majority of heavy distillates and
residual fuels cracked in refineries goes into the production
of gasoline. Crude oils that initially yield only 15 to 20
percent gasoline-range products through distillation can
yield 60 to 70 percent gasoline when subjected to cracking.
There are basically three cracking processes: thermal
cracking, catalytic cracking and hydrocracking. Thermal
cracking is the oldest of the three, and simply heats the
large hydrocarbon molecule to temperatures exceeding 1/50
C. At one time, thermal cracking was widely used to
improve the octane number of naphthas and to produce
gasoline and gas oil from heavy fractions. However, be-
cause thermal cracking of heavy distillates for gasoline
production produces substantial quantities of less valuable
gases and low-quality gas oils, the process has largely fallen
out of use.
About forty years ago, catalysts were introduced into
the cracking process to produce a higher quality gasoline.
Catalysts enable cracking to take place at lower tempera-
tures and yield a heavier, more valuable gas as well. Higher
volumes of C3 and Cti products (propane and propylene; and
butane, butene, and butadiene) are produced, offsetting
lower volumes of methane and ethane. Catalytically-
cracked gasolines contain more branched-chain hydrocar-
bons, have higher yields, and are generally superior to
thermally-cracked gasolines. As a consequence, most re-
fineries that make gasoline from heavy distillates and gas
oil use catalytic crackers.
The major problem with catalytic cracking is that the
catalyst quickly becomes contaminated with coke deposits.
Spent catalysts must be continually separated and regenera-
ted.
Fuels Refining Page 55

Hydrocracking is a process designed to increase the

yields of high-value gasoline components, usuaJJy at the
expense of the gas-oil fraction. Hydrocracking involves
cracking in the presence of both a catalyst and hydrogen
gas. In thermal cracking, olefins (which have a lower
hydrogen/carbon ratio than paraffins) are produced and in
catalytic cracking, olefins are produced and carbon elimina-
ted by deposition on the catalyst. In hydrocracking, most of
the olefins that are produced immediately combine with
hydrogen to form short branched-chain paraffins.
The cracking process is typicaJJy very flexible and can
produce high yields of either gasoline or gas oil from the
heavier crude-oil fractions. Tesoro Alaska has recently
added a hydrocracker to its Kenai refinery in order to obtain
an I !-percent increase in the yield of gasoline and middle
distillates from each barrel of crude oil processed.
Reforming. Catalytic reforming, like cracking, is one
of the most important processes in the production of gaso-
line. The process typically uses straight-run naphtha as feed
and alters the chemical composition of the hydrocarbons by
removing hydrogen. Major changes in the composition of
the naphtha include conversion of:
* paraffins to isoparaffins
* paraffins to naphthenes
* naphthenes to aromatics
Sometimes paraffins, naphthenes, or side chains break
up in the reformer to form butanes and lighter gases, but
the principal object of reforming is to raise the octane
number of the gasoline. Aromatics have higher octane
numbers than paraffins and naphthenes; long-chain paraffins
have low octane numbers.
An ideal catalyst for reforming gasoline would convert
the Jong-chain hydrocarbon molecules in the naphtha feed to
aromatics or branched-chain paraffins. Platinum catalysts
appear to be the most selective in achieving this outcome
and also, the most active in speeding the rate of reaction.
They are also the most expensive. Other dehydration and
Page 56 Fuels Refining

reforming catalysts include molybdena, chromia, and cobalt

molybdate.
The main product from a reformer is called "refor-
mate". The butanes and lighter gases released in the
process are taken off overhead and used as fuel or processed
elsewhere in the refinery. Hydrogen is also an important
reformer byproduct that can be used in other parts of the
refinery, for desulphurisation ("hydrotreating"), or for
hydrocracking.
Polymerization and alkylation. When refiners pass
crude oil through a catalytic cracker, the lighter olefins
(butylenes and propylenes) produced are too unstable to stay
dissolved in the gasoline blends. Polymerization and alkyl-
ation were invented to combine the smaller hydrocarbon
molecules into larger ones. Polymerization combines identi-
cal molecules, while alkylation combines different types of
molecules. Thus, butenes (CtiHs) are polymerized to
octenes (C6H16); similarly propylene (C3H6) becomes
hexene (C6HJ2). Propylene and butene combine through
alkylation to form heptene.
The use of alkylation has grown at the expense of
polymerization, primarily because alkylation yields more
product from the same quantity of olefin feedstock and the
resulting alkylate has superior gasoline-blending qualities.
Alkylation is also used to manufacture petrochemical de-
rivatives. For example, benzene and ethylene may be
combined to form ethylbenzene, which in turn, is used to
make styrene and synthetic rubber.
Isomerization. lsomerization involves changing the
structure of a hydrocarbon to yield a different, more
valuable isomer. In most cases, normal paraffins are
changed with the aid of a catalyst to branched-chain paraf-
fins. An original application of isomerzation was the
conversion of normal butane to isobutane for use as an
alkylation feedstock. However, with increased yields of
isobutane from reforming operations, this application is
limited. Most isomerization units now convert low octane-
rated pentane and hexane into their high-octane isomers.
Fuels Refining Page 57

Hydrotreating. As petroleum fractions move through

a refinery, impurities in the crude oil can have a detri-
mental effect on equipment, catalysts, and quality of the
finished product. Hydrotreating removes most contaminants
by mixing hydrogen with the crude-oil fractions and then
heating the mixture under high temperature and pressure in
the presence of a catalyst. Several reactions can take
place:
Hydrogen combines with sulfur atoms to form
hydrogen sulfide (H2S).
Some nitrogen compounds are converted to am-
monia.
Metals entrained in the oil are deposited on the
catalyst.
Some of the olefins, aromatics, or naphthenes
become hydrogen-saturated, and some cracking takes
place, causing the creation of some methane, ethane,
propane, and butane.
Hydrotreating is, therefore, used both to remove im-
purities and to alter the composition and characteristics of
refined products. Gasoline may be treated in order to
hydrogenate olefins and diolefins in order to reduce gum
formation.
Reformer feedstocks and other feedstocks may be
treated to remove sulfur, nitrogen, and other impurities that
could otherwise "poison" and deactivate the catalysts.
Kerosene and lube oils may be treated to reduce both sulfur
and the proportion of aromatics. Many refineries have also
added hydrotreating units to desulfurize residual fuels in
order to meet environmental specifications.
Refinery Technology and Design.
Refinery design and the choice of refinery processes
depend upon several factors, including the type of crude oil
available as feedstock, the desired product slate, product
quality requirements, and economic considerations such as
relative crude-oil prices, product values, availability of
electricity and water, air and water emissions standards,
Page 58 Fuels Refining

and the cost of land, equipment, and construction labor, and

the owners' access to capital.
Complexity of product slates adds to the complexity
of a refinery and thus to its fixed and variable costs, as does
a mismatch between the grade and quality of available
feedstocks and the desired product slate. Thus, refinery
capital and operating costs tend to be higher on the West
Coast of the United States, where product slates emphasize
lighter products; air-quality standards are more critical, and
the typical crude oil is, unfortunately, of lower gravity and
higher sulfur content than elsewhere in the United States.
A typical U.S. refinery that produces more than one
grade of gasoline and several kinds of middle distillate
products is likely to have a fairly complex array of proces-
ses, as indicated by the flow chart in Figure 2. This
complexity has evolved over a period of many decades, in
response to a growing diversity of petroleum-product de-
mand and ever more critical product specifications gene-
rated by more sophisticated fuel-using equipment.
Although "downstream" process complexity, pressure
and temperature controls, and other dimensions of refinery
technology have advanced continually over the years, crude-
oil distillation remains the heart of the refining business,
and its technology remains much as it was decades ago. All
refining operations begin with the separation of crude oil
into various fractions with different boiling-point ranges.
This is where the similarity ends. Some small refineries,
like Mapco's North Pole plant and the Chevron Kenai refin-
ery are simple "topping plants", that sell a narrow range of
straight-run distillates as final products, exactly like the
typical refinery of one hundred years ago. The essential
difference is only that the "top" and "bottom" ends of the
crude-oil barrel (the lightest and heaviest fractions) are no
longer discarded, but are now sent on to more complex
refineries for further processing or sold for electric-utility
boiler fuel or ship's bunker-oil use, where product quality is
not a critical factor.
 'Tl
PHODtJCTION 1''1tot_ OJ\
c

JJ Pf<OPANE PROPYlENE Pf\Of'MJ~ ;LPG,

"t;;"

J
GASE fHOM
TOPPlNG "NO -
H
BUTANE - BUfYlE'lt
ALKYLATION "':;"....,
[
Cf\1'CKING UNI! S
CRUC;E OIL
STORAGE L...I
GAS RECOVERY AND
L..J
I '----- T """''"'M'"''""
(/l
" "
~

Oil

-[
FEED PREPARATION
9UTANE 1lPG)
~ euTANE ..>.NO A,00Jf1VE$ IWIA1".lN GA~CUNE
'O
-------- ;:;;

~
AROMATIC TOLUENE
REFORMING EXTRACTION
SENZ ENE il~NZHJ~ !TO PHROCEMICALSI
DESALTING jPLATFORMER)

.... eur ..... e

TETRAHHYL lEAO, TCP ,.,.0 OTHER AOOIT!VES

c.o.,;oc.1 .. f Ll
t:E
LIGHI ANO HEAVY NAPHTH" CASINGHEAO ETC
1Srf!Al0'1I RUN GA$0l tNEI l I MO!C)f\ ,;. ... ,_.,,. INE

KEROSINE
+ t Kt/WSNt "NO .JH ftJE 'Tl
I ......... I ~
T
fl
01snu "HS ('")Oil

~
HfAllNGO.lMJ(JQIE~tL cLIC\

I '"'"' CRACKING s wie

0
:;w!:ElENIN<J AND
::r Ei
~"
a... "'
GASOLINE
C."-~ UH
iC"TAlYTICANO 8LENl)11l!; Rt><[JU"l l'!Jl, '-"'
CRUDE OIL
DI"- TILLA TlON THEf<M"lJ
iTOPPINGi

'"
J-(!:"
CO~E

TOP<'lll C.H<JOE COKING I OllY

w ... x
WAX SWEATING
OE011 to
W"x
CLAY
TREATING
PAR~f-fl'< .,.,,., ~
~
_J
~
co .. e
R"W PTR0l l<!M
WU
OISfiUAlf SOLVENT "001T11ES '<
PUfHflEO OIL l~~W~AA~,~~\.
HHANEI
LU8t:: Otl
BLENDING
"1 LUf>"<ICAflM" 011

VACUUM
DIS f ILLA TION HA'IY I'- I-
I
TRCATn--1r,
;;OlV(N! l OEWA~ING
. O(WAXELJ <.lit CLAY TREATING.
'V
~
OSr1u ... , E SOLV~Nf

GAEA$( Gf<t~~f

1: IMPl!f<lfl~~
MANUFACTtJflE
"
ro CRAC:ol"'Li
"'
"JEIE
,-,~~IO<lM
SOLVENT
EXTRACTION
1>URfUf<A<
"'
OUO-:S(ll)

Simplified Flow Chart of a Refinery.

Page 60 Fuels Refining

Other more complex refineries like Tesoro's Kenai

plant process the straight-run distillation products much
further and crack much of the heavier fractions into more
valuable refined products. The state of the art today is
represented by complex refineries like the one depicted in
Figure 2 and that which Charter Oil contemplated for
Valdez, in which the entire crude-oil barrel would have been
processed into gasoline, middle distillates, and petrochemi-
cals.
Forces for Change
The OPEC price revolution. The recent "energy crisis"
began in 1973-7 4 with the Arab oil embargo. The embargo
came (I) just at the peak of an unprecedented world-wide
economic boom that had stretched global oil-producing
capacity to its limit and (2) just as U. S. crude oil production
had reached full capacity and peaked out. The Organization
of Petroleum Exporting Countries (OPEC) seized upon the
shortage caused by the embargo to increase world crude-oil
prices more than fourfold. A second supply pinch, and a
further threefold price increase came in 1979-80, when the
Iranian revolution and the subsequent war between Iran and
Iraq deeply curtailed production in both countries, which
were the world's number-two and number-three exporters
respectively.
Higher oil prices and the fear of future supply inter-
ruptions have created strong incentives for consumers to
conserve energy or to change fuels (switching from oil to
coal, for example), and for the oil industry to explore for
petroleum outside the OPEC countries, and for various
parties to develop alternative energy sources. The full
adjustment of industrial economies to higher oil-price levels
and supply insecurity would have been gradual under any
circumstance. Fuel-use patterns are embodied in buildings,
appliances, transportation equipment, and industrial proces-
ses that take several years to wear out, become obsolete, or
in many cases, even to become economic to retrofit. It also
takes several years to mobilize and carry out successful oil
and gas exploration programs or to design and build substi-
Fuels Refining Page 61

tute-fuel production facilities (for shale oil extraction,

synthetic fuels, etc.).
In the United States, the adjustment was delayed even
further because the initial policy response to the events of
1973-74 was to impose price controls on domestically-pro-
duced oil in order to shelter consumers as much and for as
long as possible from the impact of rising OPEC prices. The
average inflation-adjusted retail price of gasoline, for ex-
ample, was only 10 percent higher in 1978 than it was in
1973. Not only did crude-oil price controls maintain the
level of U.S. petroleum-product consumption higher than it
otherwise would have been, but the crude-oil price-averag-
ing mechanism (the "entitlements" system) that went with it
effectively subsidized the domestic refining sector and
protected it from foreign competition.
The temporary fool's paradise that petroleum price
controls and allocation created for consumers and refiners
alike is now over. As a result, five interrelated factors are
now pressing the U.S. petroleum-refining industry --- (I) an
overall decline in petroleum products consumption, (2) a
shift in the mix of products demanded, (3) a worsening of
the average quality of crude-oil supplies, (ti) a less secure
crude oil supply, and, of course, (5) higher crude-oil prices.
Declining consumption. Total U. S. consumption of
petroleum products fell from I 8.8 million barrels per day
(MMB/d) in 1978 to about 17.0 MMB/d in 1980, and in April
1981 was less than 16.0 MMB/d. Declining product sales
have resulted in redundant refining, storage, transportation,
and distribution capacity. Consumer resistance to higher
gasoline and fuel oil prices has joined with higher crude-oil
costs to create an intense profit squeeze on refiners,
distributors, and retailers alike.
Changing market requirements. Higher oil pr ices and
federal regulations have combined to create a trend away
from both the lighter and the heavier petroleum products
(e.g., gasoline and residual oil) toward middle distillates
(e.g., jet fuel, diesel fuel, and No. 2 heating oil) and from
leaded to unleaded gasoline. Higher crude-oil prices have
tended to shift petroleum product demand away from heavy
Page 62 Fuels Refining

fuel oil, which can rapidly be supplanted by coal or natural

gas in most of its uses. At the same time, voluntary
conservation and more fuel-efficient cars (with some help
from the economic recession) have already reduced overall
U. S. gasoline consumption by more than 15 percent below
its 1978 peak. The National Petroleum Council (NPC),
nevertheless, forecasts demand for high-octane unleaded
gasoline to double by 1990. Consumption of gas oil and
naphtha as petrochemical feedstocks is also expected to
increase as demand continues to grow for synthetic textiles,
fertilizers, plastics, and other chemical products.
Declining crude oil quality. Light (high-gasoline) and
I
sweet (low-sulfur) crude oils have long been preferred refin-
ery feedstocks, particularly in North America, where motor
fuels are an exceptionally large part of total petroleum
demand and where air quality became a major concern
earlier than in Europe and East Asia. Fortunately, the grade
and quality of North American crude oils (other than in
California) has tended to be well suited to the mix of
domestic product demand.
Throughout the I 970's, however, crude-oil production
from historical domestic sources declined. As a result,
price premiums for light, sweet crudes have widened, and
U.S. refiners have had to turn increasingly to heavier,
higher-sulfur crude oil supplies, both domestic and imported.
According to the National Petroleum Council (NPC) study
cited earlier, 80 percent of the world's remaining crude-oil
reserves are high in sulfur, but only 116 percent of the raw
material run in U. S. refineries in 1978 came from high-
sulfur stocks. NPC forecasts that high-sulfur crudes will
increase their share in total feedstocks to between 55 and
59 percent in 1990. Most industry analysts appear to agree
that the trend toward heavier, higher-sulfur feedstocks will
continue, and will require refiners to make major modifica-
tions in existing plants, above and in addition to those
investments needed to deal with the shifting demand mix.
Concern for feedstock security. For several decades
before 1973, a large excess of oil-producing capacity existed
in Texas, Louisiana, and other states, where production was
Fuels Refining Page 63

controlled and allocated by State oil-conservation authori-

ties. Excess capacity in the oil-producing nations of the
Middle East and the Caribbean was even greater, and the
vast bulk of this capacity was controlled by the major
multi-national (mainly U. S.-based) oil companies. As a
result, many North American refiners were self-sufficient in
crude oil or nearly so.
Domestic and world crude-oil markets were, there-
fore, normally buyers' markets, and access to crude oil was
not a major concern for most refiners. The upheavals of the
1970's, however, made security of crude-oil supply of para-
mount interest to refiners as well as governments. First,
U.S. domestic production peaked in 1970 and declined
throughout the decade, while consumption continued to
climb until 1978, leading to an ever-greater dependency on
imported oil. At the same time, foreign oil-producing
countries were in the process of nationalizing the conces-
sions of the multinational companies. The combined effect
of these two trends was to place almost every refiner in
North America in a position of depending on other domestic
or foreign producers for a large part of their refinery
feedstocks.
Because of the two major interruptions of Middle
Eastern production that occurred during the I 970's, markets
for both foreign and domestic crude oil became dominated
by political considerations. Not only has the world's overall
supply become vulnerable to curtailment at the whim of a
handful of governments or perhaps of a handful of terrorists,
but even in the absence of an overall supply crisis, the price
that different refiners have to pay for crude oil of a given
grade and quality might vary by several dollars per barrel,
depending on the refiner's relationship with the Saudi Arab-
ian or other OPEC producer governments, or (at least until
January 1981) on the company's regulatory status under U.
S. oil price and allocation rules.
In the "seller's market" that prevailed during the
l 970's, an assured supply of crude oil seemed to be very
important to the long-term viability of existing refineries,
an important precondition for financing the construction of
Page 6/t Fuels Refining

any new refinery, and an absoultely necessary condition for

financing any independent refinery. Would-be independent
refiners, like the various groups that promoted the Alpetco
project at Valdez, seemed to center their entire investment
strategy on the search for assured crude-oil supplies, on the
apparent theory that such a supply was not on! y necessary
but sufficient for project success.
As a result, there have consistently been companies
willing to pay a premium over the benchmark price applic-
able to a given kind of crude oil, like the official Saudi
government price or Alaska's "Exhibit B" price (the weight-
ed-average of prices posted by the North Slope producers),
in order to secure captive reserves, long-term purchase
contracts, or long-term allocations by governments.
Any large new source of secure domestic crude oil
that was not yet under the control of a major refiner thus
became a particularly attractive property, and was eagerly
sought out by refiners or by speculators confident that
control over crude oil would either make them into refiners
or allow them to capture part of the premium that refiners
would pay to be assigned the right to that crude oil.
In this situation, Alaska's right under its oil and gas
lease contracts to take oil royalties either in money or in
kind gave the State two special choices for using its North
Slope royalty crude. This option could be used, on the one
hand, to attract refinery and petrochemical investment in
Alaska, seemingly even without any discount on royalty-oil
or gas feedstocks below the "in-value" price --- the amount
the State would have received if it took its royalties in cash
from the North Slope producers. Alternatively, royalty oil
taken in kind could be sold on long-term contract to existing
Alaska refiners or to refiners elsewhere at a premium above
its in-value price.
An example of the first strategy was the State's
contract with a series of groups --- most recently a Charter
Oil subsidiary (the Alaska Oil Company) --- to sell JOO mb/d
of North Slope royalty oil at the "Exhibit B" price, condi-
tional upon the company building a worldscale refinery in
Alaska. The second strategy is illustrated by the State's
Fuels Refining Page 65

1980 auction of North Slope royalty oil in approximately 5

mb/d lots for a one-year term beginning in July 1981. The
high bidders in this auction offered premiums ranging up to
almost $3.00 per barrel above the price the state would have
received if it had left the royalty oil under control of the
North Slope producers and taken payment "in value" ---
that is, in cash rather than oil.
Costlier crude oil. The average price U.S. refiners
paid for crude oil increased more than seven-fold, from an
average of $4-.11 per barrel in 1973 to about $36.00 per
barrel in early 1981. Because crude-oil costs are the major
part of the wholesale price of petroleum products, large
consumer-price increases were therefore inevitable. In the
absence of government price controls, the rise in retail
prices would have led to sharply curtailed consumption of
petroleum products, refinery and distributor margins would
have fallen nearly to zero, and there would have been little
incentive for anyone to think of investing in new refinery
capacity.
Until the beginning of 1981, however, ceilings on the
domestic price of crude oil permitted U.S. petroleum con-
sumption to keep growing through 1978. Despite the lip
service that federal policy paid to energy conservation, the
apparent demand for new refinery facilities in the United
States continued to grow apace. Price controls, moreover,
were augmented by an elaborate system of "entitlements",
under which refiners who processed price-controlled domes-
tic oil subsidized refiners who depended on imported crude-
oil and major companies subsidized small refiners. Because
U.S. refiners could buy crude oil at lower average prices
than in any other advanced country except Canada, entitle-
ments effectively insulated domestic refiners from competi-
tion with products refined abroad.
In addition, refiners and distributors were generally
able to pass through the crude-oil price increases that the
system did permit, and even to increase their markups,
because domestic product-demand remained strong at the
same time that domestic refiners were sheltered from
worldwide competition. The strong profit outlook that this
Page 66 Fuels Refining

situation generated combined with the subsidy element in

the entitlements system to encourage the oil industry to
invest in both "grass-roots" (entirely new) refineries and in
the expansion or retrofitting of existing refineries.
Finally, and rather amazingly in retrospect, almost all
of the concerned parties in industry and government seem to
have expected these market conditions to continue forever.
Throughout the I 970's, oil-company trade associations, the
Department of Energy, and both liberal and conservative
members of Congress deplored the growing "shortage" of
refinery capacity in the United States (which each of them
tended to blame on different parts of the federal regulatory
apparatus), and sponsored legislation to create new incen-
tives for domestic refinery investment.
The most important effect, for the purposes of our
discussion, was the way in which the conditions we have
described did, in fact, encourage industry to plan new
domestic grass-roots refineries. One such proposal was, of
course, the Alpetco project at Valdez.
Alpetco and other U.S. refinery-construction projects
planned in the late I 970's rested on the assumption that the
I 980's, like the l 970's, would be another decade of (l)
growing petroleum-products consumption and (2) sellers'
markets for crude oil. If these two assumptions had been
valid, they would have meant that an assured supply of
crude oil almost guaranteed the profitability of any new
refinery. The absence of either condition, however, jeo-
pardizes all current plans for domestic grass-roots refinery
construction and also casts a shadow over many of the
planned expansions and retrofits of existing refineries.

Outlook for the l 980's.

It is likely that the current (August 1981) oil "glut"

foreshadows an entirely different kind of petroleum market
in the l 980's from that which prevailed in the previous
decade. World oil consumption may well have peaked-out in
1978, and world energy prices prices may have reached their
long-term summit at the beginning of 1981, at least in
Fuels Refining Page 67

constant-dollar terms. The buyers' market that exists today

could even, conceivably, become a rout in which OPEC
prices collapse nearly as fast as they rose. More likely,
prices will remain well above 1973 and even 1978 levels, but
neither refiners nor governments will any longer seem
desperate to obtain crude oil at almost any cost.
Other scenarios are also plausible. The current glut
depends both on falling world consumption and on the
decision of the Saudi Arabian government to maintain high
production levels in order to assert its own control over
OPEC. Saudi policy could change radically overnight, the
present regime might be overthrown, or a wider war could
sharply curtail exports from the entire Middle East. If any
or all of these events came about, we would once more see
world oil prices soar, until a new equilibrium (and a new oil
glut) was established at the new price level.
If, as is more likely, oil is in fact plentiful enough
during the l 980's to exert a continuing downward pressure
on world oil prices, the consequences for oil-producing
regions like Alaska would, of course, be profound. Not only
would their oil-sales revenue be far lower than they now
anticipate, but the attraction of long-term feedstock-supply
security would no longer tend to override the transport and
construction-cost handicaps of frontier regions as a site for
worldscale refining operations.
Ironically, however, the resumption of real-price in-
creases for crude oil would not improve the generally-dim
outlook for new refinery construction in areas like Alaska,
because higher prices would cause domestic and world oil
consumption to shrink even more drastically. The present
excess of refinery capacity in the United States and else-
where would continue to grow, probably assuring that no
new export refinery anywhere --- and certainly no such
refinery in a comparatively high-cost environment ---
would be profitable.
The impact of the new situation on the outlook to
refinery modifications is less obvious, but even the drive to
upgrade existing Lower-IJ.8 refineries in order to produce a
Page 68 Fuels Refining

different product mix, or to run a different mix of crude-oil

feedstocks may be a movement whose time has passed.
One way of viewing the impact of declining consump-
tion on the need to modify existing refineries is to assume
that refiners generally prefer to run lower-sulfur, higher-
gravity feedstocks because they are cheaper to process, but
that the refining industry is facing a steady decline in the
physical availablity of such crude oils. However, a one-
percent annual decline in overall petroleum-product con-
sumption, or even a one-percentage-point reduction in the
expected rate of consumption growth, would more than
offset the roughly one-percent annual decline expected in
the supply of higher-quality crude oils.
At lower overall consumption levels, therefore, the
need to run inferior feedstocks would be considerably less
than expected. Moreover, with refineries operating at less
than 70 percent of capacity in North America, and at even
lower utilization rates elsewhere, the flexibility of the
refining sector as a whole would be greatly enhanced. As a
result, the ability to process heavy, high-sulfur crudes in
existing equipment would improve at the same time the
need to do so would be far less pressing. Circumstantial
evidence of such a tendency has already appeared this year,
in the form of lower world-market price premiums on light,
low-sulfur crudes --- a significant reversal of the trend that
dominated the l 970's.
The next few years may well be a period in which most
refineries are merely holding on with their current design,
with the least adjustable facilities being shut down. It is
important to note that the most definitive studies of refin-
ery flexibility were completed before the latter half of 1980
--- when it first became impossible to ignore the powerfully
depressing effect on oil consumption of the 1978-79 round of
crude-oil pr ice increases.
 CHAPTER 5
PETROCHEMICALS

Introduction

The manufacture of most organic chemicals relies on

the general principles of chemistry that we summarized in
Chapter 3 and on many of the same processes that fuels
refineries use to transform hydrocarbons that we outlined in
Chapter ti. The petrochemical industry's final product slate
is far more varied than that of the refining sector, however,
serving a global market with more than fifty thousand
different chemicals used in the production of food, clothing,
building materials, machinery, medicines, and many other
goods.
Because petrochemical products appear in almost
every aspect of daily life, they have become regarded as
necessities, but they also occasionally become the focus of
public consternation, as in the well-known controversies
over DDT, urea-formaldehyde insulation, and TRIS. For this
reason the development of either a new petrochemical
product or a new petrochemical-industry facility can evoke
both favorable and unfavorable public interest.
Table 9 shows that fabrication of consumer products
from petrochemicals completes a complex chain of proces-
ses stretching back to production of primary petrochemicals
from natural gas, ethane, LPG's, naphtha, and gas oil. Any
intelligent consideration of the benefits and costs of petro-
chemical development in Alaska, including health and safety
risks, requires some familiarity with these processes and
their corporate setting.
Chemical Industry Structure
The boundaries of the petrochemical industry are
rather fuzzy. On the "upstream" end, they blend into the
petroleum refining sector, which furnishes a major share of
petrochemical feedstocks; "downstream," it is often impos-
sible to draw a clear line between petrochemicals manu-
facturing and other organic chemistry-based industries such
as plastics, synthetic fibers, agricultural chemicals, paints
and resins, and pharmaceuticals.
Page 70 Petrochemicals

Table 9
Petrochemical Processes and Market Sectors

Chemical Primary Petro- Major Major

Feed- Petro- chemical Consuming End-
stocks chemicals Products Industries Uses

Naphtha Ethylene Plasticizers Fabricated Packaging,

and Plastics Construe.,
Gas Oil ProQylene Dyestuffs, Materials,
from Pigments Textile House-
Crude Oil Butadiene Products wares;
Industrial Furniture
Ethane, Benzene Organic Soaps &
Propane & Chemicals Detergents Apparel,
Butanes Q-Xylene Tire Cord
from Solvents Rubber
Crude Oil Ammonia Products Industrial,
Rubber Household
Ethane, Methanol Processing Pharma- Cleaners
Propane & Chemicals ceuticals
Butanes Tires.
from Surf- Agri- Belting,
Natural actants cultural Hose
Gas Chemicals
Plastic Prescrip-
Natural Resins tion &
Gas Patent
Synthetic Medicines
Rubber
Feed &
Medicinals Fertilizers

Nitrogen
Fertilizers

Pesticides
Petrochemicals Page 71

The chemical industry is complex, no matter how

narrowly its boundaries are drawn, and it is highly inter-
national. Petrochemicals are produced by (1) firms that
specialize in the chemical business, (2) integrated oil com-
panies, and (3) government enterprises and joint ventures
between governments and international chemical or oil
companies.
Four of the world's twelve largest chemical companies
--- DuPont, Dow, Union Carbide, and Monsanto --- are
headquartered in the United States, and each of these
companies is among the 50 largest industrial corporations in
the country, with total sales of more than $4- billion in 1980.
The chemical companies themselves generate many
captive product streams for which no public sales occur.
The chemical industry is its own best customer, particularly
for the primary petrochemicals and their first derivatives.
Dow Chemical Company and Union Carbide, for example,
produce significant volumes of ethylene oxide, which serves
as an intermediate product for making antifreeze, deter-
gents, and a host of second- and higher-order derivatives. A
large part of this output is used by the companies them-
selves, usually within the same plant or complex, and a
significant amount is sold to other chemical companies, but
very little ethylene oxide is marketed outside the chemical
industry itself. The same pattern exists for propylene,
ethylene dichloride, and a number of other primary petro-
chemicals and derivatives.
Most large integrated oil companies also manufacture
chemicals; the worldwide chemical sales of Shell and Exxon,
for example, would rank them among the top dozen chemi-
cal producers. The large-scale entry of the major oil
companies into the chemical industry reflects the compara-
tive advantage that control over hydrocarbon feedstock
supplies gives them in relation to independent chemical
producers. "Forward" or "downstream" integration by oil
companies also stems from their desire to obtain assured
markets for their future crude oil, natural gas, NGL's, and
refinery production. The recent declines in gasoline and
Page 72 Petrochemicals

fuel-oil consumption in the U.S. and elsewhere are giving

refiners a special incentive to treat petrochemicals produc-
tion as a potential outlet for surplus naphtha and gas oil.

Table 10
Worldwide Sales of Petrochemicals and Plastics

1980 Sales
Company ($ millions)
Royal Dutch/Shelli 7,633
Exxonl 6,963
Dow Chemica!2 6,882
!CI 4' 105
Union Carbidel,3 3,665
Montedison 3,594
Hoechst4 3,476
BPl 3,394
Veba 3,264
Bayer5 2,943

Notes: !) Total chemical business, exclu-

ding inter-affiliate transfers.
2) Organic chemicals, hydrocar-
bons, and plastics.
3) Chemicals and plastics.
4) Organic chemicals, plastics,
waxes, and resins.
5) Plastics, coatings, and polyur-
ethanes.
Source: Mike Hyde's Chemical Insight (No.228)

Because vertical integration has advantages for both

the feedstock producer and the processor, the forces favor-
ing integration work in both directions: Chemical compa-
nies are also integrating "backwards" or "upstream" into
hydrocarbons production, in order to reduce uncertainty
about both the prices and availability of raw materials.
Petrochemicals Page 73

DuPont's apparently successful attempt to take over Conoco

is a dramatic illustration of this trend.
The recent growth of government-affiliated petro-
chemical ventures in petroleum-producing states and other
Third World countries reflects both a striving for industrial
diversification and the opportunity that petrochemicals
manufacturing offers some of them to utilize natural gas
and NGL's that would otherwise be flared in the oil fields.
One feature that distinguishes the government enterprises
from most private petrochemical ventures is their relative
lack of downstream activity. Chemical manufacturing in
the less-developed countries has centered on producing a
few primary petrochemicals --- first-stage olefins and
aromatics, for example, and, to a lesser extent, first deriva-
tives for sale to the chemical industry.
Petrochemical Feedstocks
The primary petrochemical feedstocks include (I)
naphtha and gas oil from crude-oil distillation; (2) ethane,
propane, and butane, mainly from natural-gas liquids (NGL's)
but also from oil refineries; (3) methane from natural-gas
wells; and (4-) synthesis gas, a carbon monoxide-hydrogen
mixture that can be produced from crude oil, natural gas, or
coal. Any primary petrochemical can ultimately be made
from any of these feedstocks, but raw-material and proces-
sing-cost differences encourage chemicals producers to
choose particular feedstocks for particular products.
Natural-gas liquids (NGL's) are the principal raw ma-
terial for ethylene manufacturing in North America, ac-
counting for about two-thirds of total ethylene production.
Ethane from NGL's is a particularly desirable feedstock not
only because its processing is relatively simple, but also
because few byproducts are generated in the process. Pro-
ducers of ethylene from gas oil, on the other hand, are faced
with a large variety of compounds formed in the cracker
which, though valuable, make production facilities more
costly and complicate products marketing.
Page 74 Petrochemicals

Naphtha and gas oil are also important feedstocks for

making olefins and for the production of aromatic hydro-
carbons --- benzene, toluene and xylene (BTX). Much of
the BTX produced in North America is, however, a bypro-
duct of gasoline upgrading (naphtha reforming) in refineries,
thereby reducing the need for special aromatics-producing
facilities.
Natural gas is the principal raw material in North
America for the production of synthesis gas which is, in
turn, the main feedstock for producing ammonia, urea,
methanol, formaldehyde, and chlorinated hydrocarbons (e.g.,
carbon tetrachloride and chloroform). Although natural gas
is the preferred feedstock, synthesis gas is also produced
from coal, oil, and vegetable matter.
Feedstock supplies. The differing availability and
prices of different feedstocks have been important in shap-
ing both the development history of the petrochemical
industry and the regional variations in its evolution.
In the United States, the dominant factors governing
the development of the petrochemical industry were cheap
natural gas and natural gas liquids and elevated production
levels for high-octane gasoline. Abundant methane and
NGL's made possible the manufacture of low-cost ammonia,
methanol, ethylene, and propylene. Demand for high-octane
gasoline made aromatic naphtha less desirable for making
gasoline, and thus benzene and xylene were available rela-
tively low opportunity costs.
In Western Europe and Japan, however, gasoline has
accounted for a smaller proportion of total petroleum
consumption. Excess European refinery capacity has made
naphtha abundant and as the petrochemical industry deve-
loped, naphtha was the main feedstock for production of
olefins and aromatics, and even ammonia and methanol.
From the end of World War II to the early 1970's, the
petrochemical industry grew rapidly and enjoyed steadily
expanding markets and relatively stable raw-materials
Petrochemicals Page 75

costs. However, even before the 1973-74 oil embargo, con-

cerns about feedstock availability were starting to emerge:
* In the United States, annual natural gas and oil
discoveries during the l 960's were far smaller than the
drawdown of reserves.
* In Europe, high standards of living increased
gasoline consumption, leading to forecasts of naphtha
shortages.
* All over the world, the new crude-oil reserves
that were being proved tended to be heavier than in
the past, causing concern over the long-range suffi-
ciency of light distillate feedstocks for the petro-
chemical industry.
* And finally, developing nations wanted to start
building their own chemical manufacturing capacity,
particularly when they owned the low-cost hydro-
carbons themselves.
These pre-1973 trends became the industry's major
concerns of the mid-to-late 1970's, dominating the planning
and development of new capacity in the chemical industry.
Future feedstock developments. The total world de-
mand for petrochemical feedstocks and the factors that will
affect feedstock choices in the 1980's and 1990's are subject
to many uncertainties, including global and national eco-
nomic growth trends, the overall world oil-supply outlook,
and specific regional circumstances. The latter include, for
example, the volume of new supplies of NGL's from the
Middle East, Alberta, and Alaska.
Raw materials "availability" and supply security domi-
nated industry planing in the l 970's, and they will remain
important considerations in the choice of feedstocks. They
may not loom as large in the investment decisions of the
l 980's and l 990's, however. The reason is the appearance of
a buyers' market for crude oil and the prospect of a buyer's
market in North America for natural gas as well.
Page 76 Petrochemicals

In these circumstances, excess refinery capacity is

providing oil companies with the flexibility and added incen-
tive to market and sell the otherwise surplus portions of the
crude-oil barrel as petrochemical feedstock. The supply of
natural gas liquids to existing U.S. plants has also remained
ample thus far and, in fact, as Saudi Arabia and other oil-
producing nations export ever-greater quantities of LPG's,
competition to find markets for LPG supplies may intensify
and seriously impinge on the marketability of relatively high
cost NGL's from areas like Alaska.
In general, therefore, the physical supply of feedstocks
does not seem to be a limiting factor for Lower-48 primary
petrochemical production, but the relative costs of various
feedstocks in the I 980's and I 990's remain uncertain, and
are one explanation of an industry-wide reticence to an-
nounce construction plans for new petrochemical facilities.
Nevertheless, oil and gas prices will remain substantially
above the levels that prevailed in the early l 970's. As a
result, costs rather than availability per se will probably
play the critical role in the selection of future chemical
feedstocks, plant locations, and processes for converting
feedstocks to derivatives and end-products.
Fuel prices and processing economics will combine to
determine which hydrocarbons are to be processed to petro-
chemicals and which are more valuable as fuels. The one
most important influence on this decision --- and one of the
greatest uncertainties --- is the relationship between natu-
ral gas prices and oil prices after the former are deregula-
ted in 1985. If Lower-48 natural-gas supplies do not expand
rapidly in response to higher prices, residential and comm-
ercial consumers could bid the price of gas substantially
above that of oil-based fuels.
In this circumstance, methane would cease to be an
attractive chemical feedstock in the United States, and the
incentive of gas producers to extract ethane from the
pipeline-gas stream would be greatly weakened. As a result,
(1) new investment in ammonia and methanol production
might shun the United States altogether, moving to Canada
Petrochemicals Page 77

or to Middle Eastern countries in order to take advantage of

natural gas that would otherwise be shut in or flared, and (2)
domestic olefins production would depend increasingly on
naphtha or gas oil as feedstock.
If No. 2 fuel oil prices rose faster than the prices of
other hydrocarbons, on the other hand, new olefins produc-
tion would move to countries with surplus LPG supplies,
while U.S. plants continued to make ethylene from both
naphtha and natural-gas liquids.
Petrochemical Product Groups.
Petrochemicals are conventionally classified according
to two features: (I) their sequence in the production
process (primary petrochemicals, intermediates or deriva-
tives, and final products), and (2) their chemical structure
(e.g., olefins, aromatics, alcohols, etc.).
Primary petrochemicals are compounds with relatively
simple molecules, made directly from hydrocarbon feed-
stocks. These compounds include ethylene, propylene, buty-
lenes and butadiene, benzene, para-xylene, ammonia, and
methanol. Most of them are relatively reactive --- as a
result of multiple carbon bonds except in the last two cases
--- and it is this quality that makes them useful for
processing into thousands of more complex chemical pro-
ducts.
Olefins. Olefins are the most important primary and
intermediate petrochemicals. They are not present in crude
oil or natural gas, but are obtained when hydrogen atoms are
removed from paraffin and isoparaffin molecules, usually by
cracking. Olefins are characterized by branched or
straight-chain hydrocarbons with double bonds between the
carbon atoms. The double bonds are less stable than single
bonds and thus the olefins will readily combine or react with
other compounds.
Ethylene is by far the most important olefin for the
manufacture of petrochemical products. A typical world-
scale ethylene plant manufactures more than one billion
Page 78 Petrochemicals

pounds of ethylene per year and in 1980, 28 billion pounds

were produced in the U.S. alone.
Ethylene is a colorless, flammable gas which, because
of its extremely low boiling point (-1550 F), cannot be
shipped long distances except in high-pressure pipelines or
very costly cryogenic (refrigerated) tankers like those used
for liquefied natural gas (LNG). In the Lower 48 and
Canada, ethylene is typically produced in separate plants
and piped to other petrochemical producers. An elaborate
pipeline system has evolved in the U.5. Gulf Coast region to
connect ethylene producers and manufacturers of ethylene
derivatives such as styrene and polyethylene.

H H H H H
C=C H-C-C=C-H
H H H

Ethylene (CzH4)

Aromatics. The "aromatic" hydrocarbons are a family

of basic chemicals --- benzene, toluene and xylenes ---
characterized by the "benzene ring" molecular structure,
which has six carbon atoms and alternately-spaced double
bonds. The group is named for the distinctive odors typical
of this chemical family.
Toluene and benzene are colorless, flammable liquids,
which, like ethylene, constitute building blocks for many
chemical intermediates. Toluene and benzene are intimate-
ly related, not only because they are produced from the
same processes, but also because the principal chemical use
for toluene is the manufacture of benzene. Benzene, in
turn, is used to make a number of products, the most
notable and important of which is styrene.
Petrochemicals Page 79

Other outlets for benzene are phenol, an intermediate

for resins; cyclohexane, an intermediate for nylon produc-
tion; dodecyl benzene for detergents; aniline for dyestuffs
and rubber additives; and maleic anhydride, a raw material
for polyester glass-fiber plastics. Toluene is used to make
plastic foams, TNT and solvents.
Xylene is available from refinery catalytic reforming
processes in great abundance, but very few of the mixed
xylenes from this source have chemical applications as yet.
The major outlets for xylenes are polyester fibers, resins,
and solvents.
Most aromatics for the U.S. petrochemical industry
are made in refineries. It is not uncommon, therefore, to
locate a petrochemical plant whose product slate includes
styrene, polystyrene plastics and synthetic rubber near a
refinery producing benzene.
Synthesis gas. The term "synthesis gas" refers to a
mixture of carbon monoxide gas (CO) and hydrogen in any
proportion. In the United States, synthesis gas is made
primarily from the steam reforming of natural gas and then
processed into three major intermediate chemicals --ammo-
nia, methanol and oxo alcohols.
Ammonia is one of the world's most important
commercially produced chemicals. It is a colorless gas
with a characteristically pungent odor and is used as
the basic raw material for many different forms of
nitrogen-containing chemical compounds. These pro-
ducts and end uses include fertilizers, refrigerants,
nitric acid, water-treatment chemicals, synthetic
plastics and fibers, animal feed, explosives, rocket
fuels, and many others.
Methanol or methyl alcohol is one of the largest-
volume organic chemicals produced synthetically. A
liquid under all atmospheric temperature-pressure
combinations, methanol can be produced from natural
 Page 80 Petrochemicals

gas, coal, or vegetable matter, and thus is an attrac-

tive subsitute for liquid petroleum fuels in both sta-
tionary and transportation uses.
Synthesis gas, under special conditions and in the
presence of a catalyst, will react with olefins to
produce alcohols. The resulting oxo alcohols do not
often find their way into consumer markets. Some are
used to make solvents; but most oxy alcohols are used
in manufacturing plasticizers that keep polyvinyl chlo-
ride and other resins soft and pliable.
Intermediates or derivatives. Because of their insta-
bility, primary petrochemical compounds are typically con-
verted at the same plant or nearby into intermediate
products or derivatives, most of which are not sold in final
consumer markets, but serve as essential inputs to further
processing operations.

Figure 3
Feedstocks, Primary Petrochemicals, and First Derivatives

Natural Chemical Primary First

Hydrocarbons Feedstocks Petrochemicals Derivatives

NATURAL Methane
GAS

atural-gas
liquids
& LPG
CRUDE
OIL
Petrochemicals Page 81

Figure 3 shows the relationships among the feedstocks

and primary petrochemicals used to produce several of the
most important petrochemical first derivatives. Chemi-
cal companies manufacture many primary and intermedi-
ate petrochemicals expressly for captive product
streams. Twenty percent or more of all organic chemicals
produced in the United States remain within the same
company for further processing or are sold to other chemi-
cal companies on long-term "take-or-pay" contracts.
For example, all of the ethylene produced in the
Alberta Gas Ethylene (AGE) No. I plant at Joffre was
committed to Dow Chemical Company prior to construc-
tion. AGE's proposed Plant No. 2 plant already has custo-
mers ready to enter into long-term purchase contracts for
the ethylene it will produce. Captive streams of primary
petrochemicals and first derivatives are thus the rule, rath-
er than the exception, and help to assure chemical compa-
nies manufacturing second- or higher-order derivatives and
fabricated products a secure supply of raw materials.
Derivative products from ethylene are of particular
interest to Alaskans because the proposed Dow-Shell petro-
chemical project features extraction of gas liquids from
Prudhoe Bay natural gas and shipment of the NGL's by
pipeline to tidewater in Southcentral Alaska. There the
ethane would be separated and made into ethylene and
ethylene derivatives and the remainder of the liquids ex-
ported by tanker, probably to other plants owned by the
same companies. Table 11 summarizes the derivative
products that Dow-Shell have mentioned for possible pro-
duction in Alaska.
Final Products
End uses for petrochemicals are numerous. Petro-
chemical intermediates are converted into fertilizers; plas-
tics; all varieties of rubber and urethanes; fibers, especially
nylon, polyesters and acrylics; paints, drugs and pharmaceu-
ticals (such as aspirin and thiamine); and detergents. Pri-
mary and intermediate petrochemicals are also key ingre-
dients in making lubricating-oil additives, pesticides,
 Page 82 Petrochemicals

Table 11
Primary Petrochemicals and Derivatives Considered
For Production in Alaska

Intermediate and
Derivative Product form End-uses
ETHYLENE:
Low-density poly- Resin sold as Film for food wrap,
ethylene (LOPE) pellets, packaged garbage bags; house-
in bags, hoppers, wares, wire & cable
or containers insulation, paper
milk-carton coatings
High-density poly- Same as LOPE Blow-molded articles,
ethylene (HOPE) injection-molded bot-
tles, pipe & films
Ethylene oxide (EO) Gas in water Intermediate for EG
solution
Ethylene glycol (EG) Liquid shipped in Antifreeze, inter-
tanks and drums mediate for polyester
fiber, film, resins

ETHYLENE PLUS CHLORINE:

E thy1 dichloride Gas: seldom shipped Intermediate
forVCM
Vinyl chloride Liquid shipped in Intermediate
monomer (VCM)* tanks for PVC
Polyvinyl Solid sold as pel- Irr iga ti on and sewer
chloride (PVC)* lets, packaged pipe, electrical con-
in bags or in bulk duit, vinyl floor
tiles, rigid sheet
packaging material

(continued)
Petrochemicals Page 83

Table 11
Primary Petrochemicals and Derivatives Considered
For Production in Alaska (continued)

Intermediate and
Derivative Product form End-uses
ETHYLENE PLUS BENZENE:
E thy !benzene Gas: seldom shipped Intermediate for
styrene monomer
Styrene monomer* Liquid shipped by Intermediate for
pipeline or in tanks polystyrene, syn-
thetic rubber
Polystyrene* Solid sold in pel- Disposable drinking
lets, sheets, and cups, resin for
blocks toys, football he!-
mets, inc.

AMMONIA:
Urea Solid, sold as Nitrogen fertilizers;
prills in bags or intermediate for
in bulk urea & melamine
resins & plastics,
explosives
Acrylonitrile * Liquid shipped in Intermediate for
drums or tanks acrylic resins and
plastics, synthetic
rubber

METHANOL:
Liquid shipped by Direct fuel use,
pipeline or in intermediate for
tanks formaldehyde

*) Not listed by Dow-Shell as proposed Alaska product.

Page 84 Petrochemicals

solvents, and much more. It is unlikely that large quantities

of intermediates manufactured in Alaska will remain in the
state for processing into final products, however, both
because Alaska will remain a relatively high-cost location
for consumer-goods manufacturing, and because it is easier
and cheaper to ship intermediates long distances than to
ship a variety of fabricated products.
Petrochemical Processes and Plant Design
Petrochemical complexes are often laid out like large
industrial parks. They can include plants that manufacture
any combination of primary, intermediate, and end-use
products. For example, some ethylene plants are single-
purpose facilities that ship a single product via pipeline to
other chemical companies for further processing. AGE's
ethane-to-ethylene plant at Joffre is such an instance.
Other petrochemical complexes are composed of a number
of largely discrete, specialized plants and laboratories that
manufacture a variety of chemicals and share common
power generation and wastewater treatment facilities.
Product slates at petrochemical complexes evolve
over time, reflecting changes in market conditions and
technology. For example, in 1959, Dow Chemical Company
of Canada purchased a 700-acre site at Fort Saskatchewan.
Initial facilities included ethylene glycol, ethanolamine,
cholorophenol, agricultural chemical and chlor-alkali plants.
Within a few years, the site had more than doubled in size to
1,450 acres, and new plants were built to manufacture
caustic soda, chlorine, ethylene dichloride, vinyl chloride
monomer and ethylene oxide/ethylene glycol. The Dow
Chemical Company facility in Midland, Michigan, a much
older facility, manufactures approximately 400 chemicals in
500 plants and laboratories.
In general, petrochemical plants are designed to attain
the cheapest manufacturing costs and as such, they are
highly "synergistic". That is, product slates and system
designs are carefully coordinated to optimize the use of
chemical by products and to use heat and power efficiently.
Petrochemicals Page 85

Figure 4
Schematic Diagram of the Ethane-to-Ethylene Plant
At Joffre, Alberta

"'"
<'
"'we
~" " ~~~
o:'.!:S
~>
wo!
i~
B~
~
"~
~,
we
OU
~~~
-o
::;c!~
:;:i ~

~
L-

I
,,,,
~~
di I
I i!a

"I'
3:;;
Ci,~

i !
~~
~i
I
;
'"
iii
{ )~
! ~
.~ '
~r~
,-l
z
iJ:;;
:: .. g

~
\1 ~
iS
o-
H;;; r - - --, ~
1 ~'Pl '
_J
5
NL-
'~ { I )
'
~
,~~~~~~~
r- -,
Pl'
L - --
a! ~---J.~~
c~ H
"' i I
'1
~: ~I! '--------------'< ,.. - - .J
I

~g
=-~
,_.,~
':; ,
Page 86 Petrochemicals

For example, exothermic (heat-generating) processes pro-

vide heat for endothermic (heat-absorbing) processes; hydro-
gen-producing processes are coupled with hydrogen-using
processes; acid wastes are stored in lagoons with basic
wastes to reduce the cost of neutralization; and plant fuel is
provided in part by unmarketable hydrocarbon by products
(e.g., methane) from various processing operations.
The Dow-Shell group would take advantage of this
type of synergism in the design of an Alaska petrochemical
complex, which might produce a variety of petrochemicals
from several feedstocks --- natural gas from Cook Inlet,
natural gas liquids from Prudhoe Bay, naphtha and light gas
oil refined from Prudhoe Bay crude oil, and possibly Healy
or Beluga coal. One distinctive feature of the petrochemi-
cal complex the Dow-Shell group contemplates for Alaska is
the participation of several large companies with already-
established markets for their respective chemical products.
If an Alaska petrochemical complex should be built by this
group, it would be patterned after an industrial park where
companies operate individual plants, but they would also
take advantage of economies of scale by sharing infrastruc-
ture and transportation facilities.
To help the reader understand how an Alaska complex
might be designed and organized, the following section
presents three examples of primary petrochemical opera-
tions.
Natural-gas liquids to ethylene and its derivatives.
Ethylene is manufactured from feedstocks that range from
ethane to heavy gas oil, depending on economic conditions.
In North America, ethylene is most economically made from
ethane. An ethane-to-ethylene plant consists primarily a
large cracker whose output is mainly ethylene with small
quantities of byproducts, mostly LPG's. A "worldscale"
ethylene plant is one with a capacity on the order of l
billion pounds per year.
Figure ti shows the basic processes for ethylene manu-
facturing at the plant operated by the Alberta Gas Ethylene
Company, Ltd., at Joffre, Alberta. These are described
below:
Petrochemicals Page 87

* Ethane feedstock is vaporized and scrubbed to

remove carbon dioxide, preheated, and sent to the
cracking heaters.
* The ethane is then cracked to yield ethylene and
byproducts. The cracked gas is cooled by direct
contact with quench water and sent on to the cracked-
gas compressor.
* The cracked gas is compressed, scrubbed with
dilute caustic to remove any traces of acid gases, and
dried to remove all traces of water.
* The dried gas is progressively chilled and partial-
ly condensed at progressively lower temperatures.
* The condensate from the chilling train is separ-
ated into its components by distillation. The conden-
sate is first fed to a demethanizer where methane
goes overhead to the fuel-gas system, and the remain-
ing components go out the bottom of the column to a
de-ethanizer.
* The bottoms from the de-ethanizer go to a
depropanizer and debutanizer, where the material is
split into C2, C3, and Ct.i fractions, which are either
used as plant fuel or sold.
The overhead from the de-ethanizer goes to an acetyl-
ene-removal system where the acetylene is converted with
hydrogen to ethylene or ethane.
The stream is dried again to remove any traces of
water and sent on to a secondary demethanizer. High purity
ethylene is taken overhead, condensed and stored for use by
derivative plants.
Figure 5 illustrates the wide range of derivatives that
can be manufactured from the primary petrochemical ethyl-
ene.
 Figure 5: Derivatives of Ethylene
'"O

_ (catalyst) -+I polyethyl-1

ene polygly-
llJ
OQ
(\)

00

+NH3
r- cols
00

___t__::.
+oxygen !ethylene +H20r1 ethylene
(catalyst) oxide glycol

I y
di- & tri-
ethylene , __ +alkali ethylene
+hypo- chloro- idvcols
ethyl- chlorousf. h drin
ene acid +alcohols
from ethylene vinyl or alkyl
refin- - - +chlorine ~ dichloride - ...... chloride phenols
ery
gases
or
crack
er
--+bromine

+hydrogen
chloride
ethylene
dibromide

ethyl
chloride
L glycol &
polyglycol
ethers
'"O
(\)

+water
(catalyst) J I ethyl
alcohol
f )I
acetalde-
h de
..+
'"'<
0
()
:r
(\)
+sulfuric- sulfuric +water
acid - esters
3
~-

-
()
llJ

,I
---T V>

+benzene --.. I
ethyl-
benzene I I styrene I poly-
stvrene
Petrochemicals Page 89

Figure 6
Typical Methanol Process

Sulfur-removal W.a5le-heat llfW

vessel Reformer boiler economizer Heal recovery Gas cooler Separator

Allemo!e location of C02 addition

r-------
1
r------,. I
I I I
I
._ _____ __,
'.

Purge lo furnace

To distilloflon

Crude mathonol
Condenser
11'1-down tank
Separator Converter Redrculator Make-up gas
Converter heat compressor
ex,honger

Source: Brownstein, Trends in Petrochemical Technology

Natural gas to methanol. Methanol is produced from

natural gas as indicated in Figure 6, and as described by the
following process steps:
* First, the natural gas feedstock is desulfurized
and the hydrocarbons are decomposed in a steam
reformer. The synthesis gas thus obtained consists
mainly of CO, C02 and H. The high-grade waste heat
is used for generating steam, and some residual heat is
dissipated to the air or cooling water.
* In the next process step, the synthesis gas is
compressed to the synthesis pressure. Methanol syn-
thesis is performed at pressures on the order of 50
Page 90 Petrochemicals

atmospheres and temperatures around 50QOF, using a

copper catalyst. The heat of the reaction is used for
generating steam, and the methanol-gas mixture is
further cooled with the aid of water and/or air,
causing the methanol to condense. The unconverted
gas is returned to the reactor.
* The resulting mixture of methanol, water, and
traces of synthesis byproducts (such as higher alcohols
and dissolved gases), is purified by distillation.
* The purified methanol is then stored ready for
transportation or further processing.
Methanol made in this way can be used directly as fuel
or it can be further processed into formaldehyde, methyl
chloride, chloroform or carbon tetrachloride. Mobil Oil has
developed a process to produce synthetic gasoline from
methanol.
Figure 7
Benzene, Toluene, and Xylenes
By Reforming and Extraction

sulfolane non-aroma tic

hydrocarbons

naphtha~- catalytic extrac- extract

~r_e_fo_r_m~e-r~ ~_,.~~t-o_r~--o stripper

xylenes toluene benzene high

extract
fraction- ._ fraction- ~ fraction- .. purity
a tor ator a tor aro- recovery
ma tics
benzene
toluene
x lenes
Petrochemicals Page 91

Naphtha to benzene. Mixtures of aromatic hydro-

carbons --- benzene, toluene, and xylenes, are produced as
coproducts or byproducts in several refinery and petro-
chemical plant processes, including cracking of ethane,
naphtha, and gas oil to olefins. Most of the aromatics
produced, however, come from catalytic naphtha reformers
that convert paraffins to cycloparaffins and cycloparaffins
to aromatics. A flow sheet for the process is presented in
Figure 7.
Because the aromatics leave the reformer in a mixture
containing other hydrocarbons of the same boiling range,
the recovery process consists of extracting the aromatics
using an organic solvent and subsequent fractionation of the
individual aromatic compounds.
Benzene is obtained from the mixture of aromatics
either by direct extraction or by the hydro-dealkylation of
toluene. In this process, fresh toluene feed is combined
with hydrogen and heated. The temperature rise resulting
from the exothermic reaction is controlled by quenching
with hydrogen-rich gas. The gas stream is drawn off,
cooled, and recycled. The liquid is stabilized to remove
light paraffins and olefins, treated, and sent to a fractiona-
tor where the benzene is separated out.
As Figure 8 suggests, benzene and the other aromatics
are important primary petrochemicals for the manufacture
of styrene, nylon, detergents, epoxy resins and more.

Petrochemical Complexes and Utility Requirements

Petrochemical complexes often combine the manufac-
ture of several primary petrochemicals and derivative oper-
ations. Figure 9 lays out the different processes contempla-
ted for Phase l and Phase II of the Dow-Shell Group project.
At this time (August, 1981), the group is considering the
following feedstocks: methane to make ammonia and urea;
ethane for ethylene and its derivatives; naphtha for ben-
zene; and coal for methanol.
 Page 92 Petrochemicals

FIGURE 8
Aromatics Derivatives

Phtha!ic
Exlttlng aromatics anhydride
Refinery
naphthas
Reforming of cyclopentanes,
eye o exanes, an para ms lsophtha!lc
acid

(minor!
Terephthalic
acid
Cyclohexane

Air

Cyclohexanol,
cydohexanone

Hydroxyl' Phosgene
amine, acid
Toluene
diisocyanate
Adipic
acid
Acid
Hexamethylene- Bisphenol
diamine Sulfonation & alkali fusion

Propylene
tetramer

Dodecyl-
Polymers ben:zene

Source: Shreve and Brink, Chemical Process Industries

Petrochemicals Page 93

Figure 9
Dow-Shell Petrochemical Project

PHASE I PHASE II

polyethyl
ene 400

ethylene
oxide&:
ethylene alpha
ethane ethylene glycol olefins
1 200 600 ,,... 400

ethyili:J- styro- poly-

benzene foam ethyl-
I 500 ene 200
PHASE II

ethylene
propane 1 200 ethylene
butanes oxide&:
pentanes ex ort ethylene

naehtha I
benzene
970
glycol
600

ethylene
coal tmethano di-
4 000 ex ort chloride
chlorine 1- 700
salt and
caustic
soda
1,150
 Page 94 Petrochemicals

The utility requirements to operate a plant of the

magnitude contemplated by Dow-Shell are substantial, as
indicated by the project's forecast demand for electricity,
steam, air, and water in Table 11.

Table ll
Projected Utility Requirements: Dow-Shell Project

Utility Phase I Phases I and II

Electricity 75 mw 245 mw
Steam 350 ton/hr 615 tons/hr
Nitrogen 2,000 SCFM 3,500 SCFM
Air 7,500 SCFM 9,000 SCFM
Potable Water 300 GPM 500 GPM
Demineralized Water 2,500 GPM 4,600 GPM
Cooling Make-Up Water 8,000 GPM 12,000 GPM

SCF M = Standard cubic feet per minute

GPM = Gallons per minute
Water use. Water supply is a very "site-specific"
consideration, which affects not only the choice of plant
location, but also plant design and operating costs. Water
serves a variety of refining and petrochemical needs inclu-
ding cooling, processing, steam generating, potable water
use and sanitation.
Relatively little water is actually consumed by re-
fineries and petrochemical plants; however, huge quantities
of water are used for cooling and condensing. In many
chemical and refining processes, the feed is heated or
vaporized to promote the desired reaction or permit the
required separation of products. The products, in turn, must
be condensed to a liquid and cooled to a safe temperature
I
for storage or product blending.
A large amount of heat is recovered by the use of heat
exchangers to transfer heat between fluids; e.g., heat con-
tained in a hot product that must be condensed or cooled is
Petrochemicals Page 95

transferred to a cooler feed stream that must be heated.

This arrangement conserves fuel and reduces cooling water
requirements. Most cooling water, moreover, is pumped
within a closed system, where it is generally not subject to
contamination and thus can be reused repeatedly.
Refineries and petrochemical plants also require large
quantities of water to generate steam for power, evapora-
tion heating, and drying. Most of the steam is condensed in
closed systems and is normally reused. Water suitable for
steam generation, however, requires extensive treatment
because as the water is evaporated, solids in the boiler
water become concentrated and can cause overheating.
Also, gases dissolved in the water or liberated from dis-
solved minerals will corrode pipes and fittings.
Much smaller quantities of water are required for
process purposes. In refineries, crude oil normally contains
salt and other matter that is removed by water-washing to
avoid corrosion or fouling of process equipment. Where
water is used to separate oil and water phases, to wash
traces of treating chemicals from product streams, or to
flush lines and other equipment, the possibilities for water
contamination are high and for water reuse, relatively low.
Finally, potable-water requirements for drinking and sanita-
tion are relatively small compared to other water uses.
The quantity and quality of water needed for a petro-
chemical complex are not absolute. Dow-Shell estimate
daily water requirements to be in the neighborhood of 21/.6
million gallons, a staggering amount considering that the
City of Kodiak, for example, uses only 10 to 12 million
gallons daily at the height of the fish-processing season. If
water is not available in these quantities at the chosen site,
however, the requirement could be reduced substantially by
the use of sea-water or air-cooled towers.
Water can be a significant cost factor where access to
the water requires major investments in wells or pipelines,
or when equipment must be installed to treat or conserve
water. The Pac-Alaska LNG plant design illustrates the
relative-cost tradeoffs. The sponsors considered piping
Page 96 Petrochemicals

water 16 miles from the Kenai River, but found that an air-
cooling system was less expensive. Because of the choice of
air-cooling, the Pac-Alaska plant, if constructed, will re-
quire only about one percent of the water now used by the
much smaller Phillips LNG plant.
Energy requirements. Both the chemical and refining
sectors are substantial users of energy for boiler and process
fuel, refrigeration, pumps and compressors, etc., in addition
to their feedstock requirements. The chemical industry
consumes more than one third of the energy used by all
manufacturing industries in the United States. Because it is
a significant cost factor, plants are carefully designed to
use energy synergistically and are often located in places
where fuel is relatively cheap.

Health and Safety Issues

Whatever economic interest Alaska may have in ex-
panding the processing of locally produced hydrocarbons
within the state, other issues have dominated public discus-
sion, and will probably continue to do so. A recent survey
by the Alaska Department of Environmental Conservation
identified the largest public concerns with respect to petro-
chemical industry development as the transportation of
chemicals, public health, air and water quality, and disposal
of hazardous wastes, while employment, population growth,
and the impact of public services seemed to be less impor-
tant.
These issues were raised during the debate over the
Alpetco proposal and will surely surface again if the Dow-
Shell group or some other entity decides that petrochemi-
cals development in Alaska is economically feasible and
moves toward the design and construction of transportation
and processing facilities.
Poisons and carcinogens. The acute toxic character of
some naturally occurring chemicals is obvious, causing
death, sickness, or readily detectible biological reactions in
organisms that come into contact with them. Evolutionary
adaptation, however, has made humans and other organisms
Petrochemicals Page 97

relatively immune to small internal doses or contact with

most chemical substances they are likely to encounter in
nature. There are a few well-known exceptions, including
poisoning from heavy metals (e.g., lead, arsenic, mercury,
and cadmium) that can accumulate in the body over many
years from natural sources, and some natural "carcinogens"
(cancer-inducing agents).
Synthetic organic chemicals create special hazards,
however. Mid-stage petrochemical derivatives tend to be
particularly active chemically (hence their usefulness as
chemical intermediates). As a result, once they leave the
controlled environment of the laboratory or chemical plant,
they readily interact with many other kinds of molecules,
including those that make up human cells.
Because they are highly reactive with other sub-
stances, moreover, these chemicals tend be absent or infre-
quent in nature, even in minute quantities, and for that
reason, evolution has had no occasion to create natural
defenses against them in humans and other organisms (by
permitting only the more resistant individuals to survive and
reproduce ).
The most pernicious of the bioactive substances may
be those that have an affinity for the genetic materials in
cell nuclei --- the proteins that control cell development
and division. Today's view of cell biology implies that just
one molecule of such a chemical coming into contact with
one DNA molecule in one cell can alter the cell's "genetic
code" and thereby initiate the production of cancerous cells
or, if the DNA is part of a reproductive cell, cause a
defective birth or an inheritable defect.
One implication of this mechanism would be that there
is no safe dose for any such substance, just as there seems
to be no safe dose for ionizing radiation (which includes the
earth's natural "background" radiation and cosmic rays, as
well as x-rays and the emissions from nuclear power plants).
Controlling human exposure to carcinogens or teratogens
(agents producing inheritable defects) without
Page 98 Petrochemicals

eliminating it will only reduce the number of deaths, illnes-

ses, or defective births caused by the substance, but can not
completely eliminate them.
Not all authorities agree with this "one-shot" view of
the way in which carcinogens and teratogens operate. Some
scholars still hold to the classical theory in which there is a
"threshold" of exposure for each substance below which
harmful effects are unlikely. This older "pharmacological"
view is, significantly, the assumption that underlies federal
harmful-substances legislation and the regulations of the
Occupational Health and Safety Administration (OSHA).
Whether or not one molecule is sufficient to initiate a
cancer or a birth defect, poisoning by synthetic chemicals
has other indisputably troublesome features, including long
lags between exposure and the first appearance of symp-
toms. One famous instance is that of DES (diethyl silbe-
strol), which appears to produce cervical cancer in the
grown daughters of women who took the drug two or three
decades ear lier.
Even without such delays in the occurance of harmful
effects from a chemical, moreover, it can be years or
decades before a sufficient statistical base accumulates to
warrant suspicion that the substance is harmful, much less
to establish the fact conclusively. The first hint might
appear, for example, only when a public health statistician
noted that the last 25 years had seen three cases of a
certain rare form of cancer among the more than seven
thousand workers who had worked in a particular plant,
while the average incidence in a national population sample
of that size would have been less than one.
It is conceivable, indeed, that powerful synthetic poi-
sons and carcinogens exist that will never be detected. A
few kilograms of a given substance vented into the atmo-
sphere or carried off in the drains and diluted throughout
the world's oceans over a period of years might increase the
worldwide incidence of cancer or, say, mongolism by tens of
thousands of cases per year. If these cases were dispersed
widely enough geographically and over time, they would be
Petrochemicals Page 99

overwhelmed statistically by the hundreds of other things

that influence the world's mortality and morbidity trends.
There are now thousands of experts engaged in testing
the effects of acute and long-term exposure to various
chemicals, but little is really known. More than fifty
thousand synthetic chemicals are currenty produced in com-
mercial quantities, and about four to six hundred new
chemicals are introduced commercially each year. Only a
tiny fraction of these substances were tested for cumulative
toxicity or carcinogenic effects before being marketed. The
chemical industry and federal regulatory agencies have been
continually expanding their testing programs for newly in-
troduced substances, but a considerably greater effort would
be necessary in order to bring our know ledge of the thou-
sands of untested but currently marketed chemicals up to
the standards that apply to new products.
The development of a petrochemical industry in
Alaska inevitably involves the production, handling, storing,
transportation, and disposal of substances that are or may
be hazardous to human life and health. Three of the chemi-
cals the Dow-Shell group contemplates producing in Alaska
are known carcinogens --- benzene, ethylene oxide, and
ethylene dichloride. Two others are suspected carcinogens
--- ethylbenzene and ethylene glycol. Moreover, most
petrochemical complexes of the type that Dow-Shell are
studying also produce vinyl chloride monomer and acrylo-
nitrile, both of which are known to cause cancer.
Since World War II, sufficient evidence has accumul-
ated to establish that some aromatic hydrocarbons produced
in refineries and several aromatic products of petrochemical
plants, pose health hazards when a major spill or accident
occurs or when people are repeatedly exposed or receive
prolonged contact with even minute amounts.
Prolonged exposure to high concentration of benzene
is known to cause irreversible damage to the bone marrow
where red and white cells and platelets are formed. Ben-
zene exposure can cause aplastic anemia, a form of leu-
kemia, chromosome damage in white blood cells, and acute
Page 100 Petrochemicals

myleogenous leukemia. Benzene is also a central-nervous-

system depressant.
The federally regulated occupational exposure level
for benzene is now 10 parts per million (ppm), averaged over
an eight-hour day. OSHA regards this level of exposure to
be too high and recommends a standard of 1 ppm, a standard
that was recently rejected by the Supreme Court on a
procedural technicality in a 5-ti decision.
These standards apply principally to refinery and
chemical plant workers, but we know almost nothing about
the health effects, if any, of the tons of benzene that are
released into the atmosphere every day when automobile
gasoline tanks are filled. Little consideration has yet been
given, moreover, to systematically measuring, much less
controlling, exposure to aromatics on the part of those who
may conceivably comprise the most numerous and severely
impacted occupational group --- filling-station attendants.
Ethylene dichloride is another chemical also in the
midst of regulatory controversy. It is a major ingredient for
manufacture of vinyl-chloride monomer (VCM --- one of
the most active known carcinogens) and appears to pose
some danger itself. The current regulated level of exposure
to ethylene dichloride is 50 ppm; however, in 1975, the
National Institute for Occupational Safety and Health
(NIOSH) recommended a revised standard of 5 ppm because
impairment of the central nervous system and increased
morbidity (especially diseases of the liver and bile ducts)
were found in workers chronically exposed to ethylene
dichloride at concentrations below 1/0 ppm and averaging up
to 15 ppm.
In addition to exposure problems within the petro-
chemical plants, there are also risks associated with the
transfer and shipment of chemicals. Again, as with expo-
sure in the workplace, the implications of and dangers posed
by a chemical spill are not precisely known. Petrochemicals
do, however, present a hazard to marine ecosystems both in
terms of an acute spill situation and chronic exposure to
small dosages. The acute toxicity of ethylbenzene to
Petrochemicals Page 101

marine organisms occurs at concentrations as low as 0.1/3

ppm, for example, but little is known about the toxicity to
marine animals from chronic exposure to lower concentra-
tions of ethylbenzene or other chemicals.
A petrochemical complex in Alaska may not by itself
pose great health, safety, or aesthetic risks. Production of
some first-stage petrochemicals such as ethylene and meth-
anol is virtually odorless and, with the sometime exception
of large quantities of water vapor and occasional flaring,
the plants are fairly safe and quite inconspicuous.
The production of benzene, ethylene dichloride, and
possibly VCM, however, presents a new dimension of risk for
Alaska industry and to the communities in which the plants
would be located. Acceptable levels of exposure are the
subject of much dispute and debate even within the respon-
sible federal agencies (OSHA and NIOSH).
The Policy Dilemmas. In summary, the hard facts
about chemical health hazards are sparse, and even where
the facts are known, their policy implications are not
obvious or even easy to think about systematically. There is
an undeniable statistical association between exposure to
aromatic hydrocarbons or VCM, for example, and the inci-
dence of certain degenerative diseases and cancer. If, as is
likely, the "one-shot" view of carcinogens is correct, there
is probably an inescapable risk of exposure and some in-
crease in the risk of cancer for anyone working or residing
in places where these products are produced, stored, trans-
ported, or used.
Society tolerates cigarettes, firearms, motorcycles,
and a host of other products whose association with death
and sickness is far more obvious than that of benzene or
VCM. Public policy toward these products is just as
controversial as are policies regarding hazardous chemicals
or nuclear power, but they are tolerated in part because
their risks are regarded as largely voluntary. The distinc-
tion is not an absolute one, as tobacco, handguns, and traffic
accidents do claim "innocent" victims and, just as an indivi-
dual can choose not to smoke, ride a motorcycle, or
Page 102 Petrochemicals

associate with people who play with guns, he or she presum-

ably can choose not to work in or live next to a petrochemi-
cal plant or nuclear reactor.
Fuzzy as the distinction is, however, it is a real one.
Regulation of the health and safety risks --- known and
unknown, real and imagined - - - is a social and political
task. The people of Alaska will have to make some hard
practical decisions on how their government will deal with
these issues.

Options for Regulation

Most States have adopted standards and regulations to
govern the handling, processing, and storage of hazardous
substances and wastes. These regulations attempt to ad-
dress certain known workplace and environmental hazards.
For example, oil contamination is a serious problem in
refining and some petrochemical operations. Hydrocarbons
can enter the wastewater system directly from a spill, leaks
from lines, vessels and valves, leaks around pump packing,
or product-sampling. Contamination can also occur when oil
and water are brought into direct contact as in crude-oil
desalting operations or product washing following chemical
treating. Remedies include special collection and segrega-
tion systems and processes for removing oil from ships'
ballasts, wastewater, plant runoff, and the like.
Disposal of spent chemicals poses another set of
environmental problems with special significance to people
who live near hazardous-waste dumps established before
there were federal regulations governing the disposal of
toxic substances. Today, because feeds tocks are expensive,
hydrocarbon byproducts are generally recovered and re-
cycled, so that new petrochemical facilities tend to gene-
rate a smaller volume of waste materials than older plants,
while those waste materials that are produced are subject to
increasingly stringent regulations regarding the sale of spent
chemicals and disposal by chemical means, incineration,
venting or flaring.
Petrochemicals Page 103

Finally, air pollution in the form of emissions and

"fugitive" (unintended) leaks, is an important concern that is
reflected in the current requirement for special permits as a
condition of refinery and petrochemical-plant construction,
and operational regulations intended to protect air quality.
Major sources of air pollution are burning of fuel in boilers
and process heaters. The greatest volume of discharge from
other plant components typically occur at peak operating
conditions, during plant upset or malfunctions, and during
the startup or shutdown of operations. Atmospheric dis-
charges of sulfur oxide, hydrogen sulfide, and mercaptans
are, of course, a major problem for refineries processing
high-sulfur crude oil.
The State of Alaska has adopted and strengthened EPA
air-quality standards, and adopted and modified EPA and
OSHA water-quality standards, waste-disposal regulations,
and occupational-safety standards. The State's leasing and
permitting processes provide vehicles for controlling dis-
charges on, into, or under State lands, and into surface and
groud water. In 1981, the legislature also passed a waste-
disposal law that authorizes State agencies to write new
standards for the handling and disposal of hazardous wastes.
The Federal regulatory machinery in Alaska is com-
paratively lean but is probably adequate to the relatively
small chemical industry that exists in Alaska today. The
prospect of large-scale petrochemical development in Alas-
ka, however, suggests the wisdom of at least investigating
and comparing additional measures that might be imple-
mented at the State level to protect human life and the
natural environment.

Prescriptive vs. Economic Remedies

There are two polar approaches to control of heal th
and safety hazards and environmental quality, and a number
of in-between measures. At one extreme are prescriptive
and proscriptive regulations, which state in categorical
terms what industry may or may not do, what facilities are
acceptable, or exactly how certain equipment is to be
designed. In a prescriptive system of regulations, remedies
Page 104 Petrochemicals

for non-compliance can be either criminal penalties (includ-

ing fines and imprisonment), civil penalties, suspension of
operation, or orders by regulatory agencies or the courts
directing specific performance. The other extreme leaves
the individual or enterprise free, at least in principle, to
decide how to operate, but relies on economic incentives in
the form of graduated fees or penalties to discourage
harmful activity, and tax-credits or other rewards for
desired behavior.
Prescriptive and proscriptive regulation. Prescriptive
regulations are of two general types, specifications and
performance standards. Traditional building codes are typi-
cal of regulation by specification, attempting to limit fire
hazards by prescribing lath-and-plaster walls, protecting
sanitation by requiring cast-iron drain pipes of a certain
diameter, and the like. Many of the Interior Department's
stipulations governing the construction of TAPS were also of
a prescriptive character. Effluent and emissions standards
that set the maximum absolute volume or maximum concen-
tration of some pollutant that may be released by a single
plant or from a single point are examples of performance
standards.
The ultimate proscriptive regulation is simply to for-
bid the activities that are deemed hazardous. One possible
approach to the real and imagined hazards arising from the
production, storage, and transportation of VCM or ethyl-
benzene, for example, would be for Alaska to ban their
manufacture within the state.
It is likely that the federal courts would hold that a
direct legal prohibition on the production of specific sub-
stances was unconstitutional, but it would not be difficult
for the State to surround the industry with so many regula-
tory restrictions and so much red tape as to make it
economically unattractive. The State of Alaska has, for all
practical purposes, done this with respect to nuclear power.
The most direct and least vulnerable approach legally,
however, would be to use the state's proprietary position and
contractual ability rather than its police powers to prevent
industrial activity it deems undesirable --- just as it uses
Petrochemicals Page 105

these powers to promote in-state processing of hydrocar-

bons, timber, etc., and preferences for employment of
Alaska residents. The State, in other words, is free not to
sell its royalty oil, gas, or NGL's; not to lease plant sites,
and not to sell sell gravel from state lands unless the party
stipulates that it will not produce, store, or transport
specified chemicals within the State.
The advantages of prescriptive regulations are their
relative and ease of enforcement. Their disadvantages are
inflexibility and insensitivity to costs. Obsolete building
codes, for example, have frequently delayed the introduc-
tion of cheaper, stronger, and safer building materials; a
categorical Federal requirement for secondary treatment of
municipal waste-water has imposed extravagant sewage-
treatment costs on many small communities (including Alas-
ka communities), where it makes no perceptible contribution
to human health or environmental quality. Yet the same
regulation allows serious water-quality hazards that could
be resolved at comparatively low costs to persist in a
number of more densely populated areas.
Economic incentives. At the other extreme from
prescriptive and proscriptive regulation are purely economic
incentives that leave design and operational details, and the
risks attendant upon them, entirely up to management. The
heart of this approach in its historical form is the right of
injured parties to sue and recover damages for loss of life,
or injury to persons or property.
Litigation. Traditional civil remedies rely on the
threat of lawsuits and expensive court awards to induce
industry to spend just about as much on health, safety, and
environmental protection as the risks of measurable (and
litigable) damage warrant. The effectiveness of litigation
as a deterrent to (as well as a remedy for) private or public
injury has been greatly enhanced in recent years by, (I) the
possibility of "class action" suits, in which large numbers of
parties claiming relatively small individual injuries can
group together to litigate, (2) the increasing tendency of
State and local governments to institute proceedings to
Page 106 Petrochemicals

recover for alleged damages to public values in cases where

it would be difficult to show or measure individual damages,
and (3) the publicity accorded to a few huge settlements and
awards of punitive damages in occupational-injury and pro-
duct-safety proceedings.
Strict liability. Traditionally, civil remedies for
injuries to health, safety, or the environment are available
only to injured parties who can prove that there was
misconduct or negligence on the part of the persons that
caused the problem. The very existence of a refinery,
tanker terminal, or petrochemical plant, however, creates a
statistically certain risk of injury or damage to someone,
sometime, even if there is no legally provable misconduct or
negligence on the part of anyone. (Suppose a wholly
unanticipated natural disaster ruptures a tank full of
poisonous gas; or suppose that a chemical which was
rigorously tested turns out to have horrible long-term
effects that no one reasonably could have been expected to
anticipate?)
Thus, for the possibility of litigation to be an adequate
remedy, legislation is necessary to make the legal liability
for certain kinds of damage "strict", or "absolute" --- not
conditional upon proof of negligence, in other words. For
example, Alaska law establishes strict liability for damages
from marine oil spills.
Individual litigation is inadequate or totally inappli-
cable, even with strict liability, wherever damage is likely
to be distributed randomly over a large and hard-to-define
population (as is often the case with carcinogens), so that
responsibility cannot clearly be assigned. It is also inade-
quate where the values to be protected are not privately
owned (as in the case of a commercial fishery stock), or
difficult or impossible to put a price on (air clarity, for
example, or the ability of an area to support a wild bird
population).
Insurance. Another problem with relying on the
judicial process to motivate safe design and operation of
industrial facilities is the cost of litigation, its overall
Petrochemicals Page 107

uncertainty, and the long time that typically elapses bet-

ween the damage and its compensation. Insurance, and
particularly insurance funds administered by an independent
party, can benefit both industry and the public by cutting
legal costs, delays, and the uncertainty of the outcome.
Insurance can be either voluntary or mandated by Jaw:
There are a number of Federal, State, and cooperative
insurance funds for clean-up after major accidents.
The Trans-Alaska Pipeline Liability Fund was the first
Congressionally created entity of its kind, receiving a fee of
five cents per barrel lifted at the Valdez terminal by the
TAPS owner companies. The purpose of the fund is to pay
legitimate claims for damages, including clean-up costs,
resulting from oil discharges between Valdez and any other
U.S. port; the Fund is liable without regard to fault for that
increment of damages in excess of $1 ti million but not in
excess of $100 million per oil-spill incident.
Insurance also has its shortcomings, however. As
many readers who have had difficulty with an auto insur-
ance claim may recall, it sometimes requires litigation to
collect benefits, even from one's own insurer. Diluting the
penalty a firm pays for a given injury also dilutes the
incentive to avoid the injury. Premiums in private insurance
programs are normally adjusted to the experience of the
individual enterprise as well as the industry, but rigorous
actuarially-based premiums seem to be the exception in
compulsory, government-sponsored no-fault insurance pro-
grams.
A difficult issue in establishing mandatory insurance
programs is whether to make them serve as a substitute for
all other private remedies, or to allow those who are injured
to retain all of the rights they would otherwise have under
civil law. Workmen's compensation laws in most States
prohibit lawsuits or any other remedy for job-related inju-
ries. The federal insurance program for nuclear power
plants (the Price-Anderson Act) also eliminates any other
recourse on the part of individuals who might be injured in a
nuclear accident. Details of the issue are beyond the scope
Page 108 Petrochemicals

of this report, but either choice may create serious inequi-

ties.
Effluent taxes and hybrid systems. There are a
variety of health, safety, and environmental regulation
techniques that are not based purely on an economic asses-
sment of risk, but neither are they purely prescriptive.
Toward the economic end of the spectrum, there is a
growing literature on the advantages of emissions and
effluent taxes. Few such programs yet exist in North
America, but they are common in Europe.
A regional air-quality board might, for example, est-
ablish a tax or penalty per kilogram of sulfur dioxide (S02)
discharged into the air. This tax would allow each operator
of an electrical generating plant or refinery to decide
whether it was cheaper to reduce emissions or to pay the
tax. The S02 tax rate could be adjusted periodically to
create just enough pressure on industry as a whole to hold
the concentration of 502 in the atmosphere below some
target level. Since those firms with the cheapest options
for reducing sulfur emissions are the ones that would choose
not to pay the tax, this approach would theoretically ach-
ieve a given improvement in air quality at a considerably
lower economic cost than prescriptive standards.
A related regulatory technique is to establish prescr ip-
ti vel y based --- even arbitrary --- standards for some
aspect of environmental quality, but to allow a "market" in
pollution rights. A regional air or water quality agency
might, for example, establish a maximum allowable rate of
discharge or percentage concentration of certain pollutants
for each plant or effluent source. A plant would be allowed
to exceed its individual pollution quotas or even to establish
a new source of pollution, however, if it could induce some
other party to reduce its output of pollutants by a compar-
able amount.
California's Air Resources Board, for example, requir-
ed Sohio to effect a net reduction in certain air pollutants in
the region, as a condition of the Board's licensing of the
tanker terminal that the company planned at Long Beach to
Petrochemicals Page 109

serve its proposed Pactex pipeline. Sohio's solution was to

pay for smokestack scrubbers for electrical generating
plants owned by the Southern California Edison Company.
The system that Alaska establishes to control the
health, safety, and environmental risks of hydrocarbons
processing will undoubtedly differ from the current mix of
prescriptive, proscriptive, and economic regulation that
exists in Federal law or in the laws of other States (or other
nations). It is in order, however, for Alaska to begin a
systematic review of these systems, their effectiveness, and
their cost-effectiveness.
 CHAPTER 6
THE ECONOMICS OF HYDROCARBONS PROCESSING
AND THE OUTLOOK FOR
REFINING AND PETROCHEMICALS IN ALASKA

Three cost factors dominate investment and location

decisions for hydrocarbons-processing facilities: (1) tran-
sportation costs, (2) feedstock costs, and (3) plant-construc-
tion costs, It is the interplay among these three factors
that determines whether export refineries or petrochemical
plants will be built in Alaska.

Hydrocarbon Transportation Economics

Transportation cost is the most powerful economic
influence on the location of refineries and petrochemical
plants, and one of the most important considerations in
c hoosing a product slate. Two fundamental axioms govern
the relationship betwe en transport costs and the choice of
transportation systems and plant location:
(1) Light hydrocarbons cost more to ship
per unit of weight or energy than heavy hydro-
carbons. A corollary of this axiom is that gases
cost more to ship than liquids.
(2) Tankers are the most efficient long-
distance transportation mode for hydrocarbons
that are liquid under atmospheric conditions,
while pipelines are the most efficent mode for
gases.
The first axiom and its corollary rest on elementary
physical principles. At a given pressure and temperature,
solids and liquids pack more matter and more energy into
the same pipeline or tanker space than gases. A cubic foot
of propane gas contains more energy than the same volume
of methane gas, and a barrel of crude oil or residual oil
contains more energy than lighter petroleum products like
gasoline or naphtha, or light chemical derivatives such as
methanol.
Page 112 Economics and Alaska Outlook

Waterborne bulk carriers. The second axiom reflects

the fact that waterborne transport is generally the cheapest
way of moving a given weight or volume of any bulk
commodity. Crude oil, in turn, is almost an ideal cargo for
large ocean-going vessels. It has just the right density ---
slightly lighter than water so that the entire hull-space can
be filled with cargo and the vessel will have a low center of
gravity, which improves its stability. A liquid at atmospher-
ic pressures and temperatures, crude oil does not require
closely-controlled conditions on board, is easy to load and
unload, and is relatively insensitive to contamination.
Shipping gases by tanker is a wholly different matter.
Vapors must be chilled and liquefied in costly facilities that
consume substantial amounts of energy as well. The lightest
hydrocarbons such as methane, ethane, and ethylene have
especially low boiling points, and vessels designed to carry
them feature expensive cryogenic (refrigerated or super-
insulated) compartments.
The heavier propane and butanes (LPG) require less
energy to liquefy, and will remain liquid at atmospheric
temperatures if confined in tanks under only modest pres-
sures. Thus, while ocean-transport costs for LPG are
substantially higher per unit of energy than for crude oil,
LPG is much less troublesome than natural gas, ethane, or
ethylene.
Gas-pipeline transportation. Pipelines are the ideal
transport mode for gases. In a pipeline, extremely high
pressures can be used to squeeze even the lightest hydro-
carbons into dense-phase fluids that contain nearly as much
energy in a given space as liquids, and these fluids can be
pumped long distances with a relatively modest loss of
energy in the form of compressor fuel.
TAPS vs. ANGTS. In Alaska, these principles can be
seen in the contrasting choices of transportation modes for
North Slope oil and gas. Before deciding to build an oil
pipeline across Alaska, the North Slope producers investiga-
ted the feasibility of a sea route directly from Prudhoe Bay
to the U.S. East Coast. The all-tanker system was rejected
Economics and Alaska Outlook Page 113

in favor of a pipeline only because of the delays that would

have been involved in perfecting ice-breaking techniques.
While an all-pipeline system across Canada would have
been the cheapest way to take Alaska oil to the Upper
Midwest, the companies finally chose TAPS because it was
the shortest land route to a year-round ice-free port (Val-
dez), from which tankers could carry crude oil for less than
$1 per barrel to any Pacific Coast port in North America or
Asia.
For Prudhoe Bay natural gas, in contrast, most parties
favored an all-pipeline route across Canada over a liquefied
natural gas (LNG) tanker system right from the beginning,
because of the latter's greater capital cost and fuel con-
sumption. Even now, if transportation of North Slope gas by
means of the proposed Alaska Highway pipeline turns out to
be so expensive that the gas cannot be marketed in the
Lower 48, the gas producers are unlikely to reconsider an
all-Alaska-pipeline/LNG system. Even the El Paso Compa-
ny, which originally sponsored the LNG concept, has aban-
doned it as uneconomic.
If an overland pipeline from the North Slope turns out
to be infeasible, the relative costs of transporting different
hydrocarbons suggest that:
(1) There may be NO alternative trans-
portation technology that is capable of moving
North Slope methane to market at an acceptable
cost.
(2) Such a system, if it does exist, is less
likely to combine a Trans-Alaska gas pipeline
with LNG tankers than to involve either ---
(a) processing the natural gas right on the
North Slope into products like methanol or
synthetic gasoline that could be shipped to
an ice-free port through the existing oil
pipeline or a new liquid-products pipeline;
or
Page 114 Economics and Alaska Outlook

(b) shipping LNG directly from the Arctic

Coast in icebreaking tankers or submarine
barges.

Transportation costs and plant location. The two

transportation axioms above can be applied directly to
decisions that govern the location of refineries and petro-
chemical plants:
(1) Petroleum refineries tend to be located
near their markets.
(2) Naphtha and gas-oil-based petrochemi-
cal plants tend to be located near refineries.
(3) Natural-gas-based petrochemical plants
tend to be located near their raw-materials
sources.
Refined petroleum products are more costly to ship
long distances than crude oil, and only partly because of
their lower energy-density. Refineries produce a variety of
products with different viscosities and vapor pressures, and
with different degrees of flammability, toxicity, etc. Indi-
vidual refinery products are therefore typically shipped in
relatively small batches and tend to require specialized
treatment to avoid loss or contamination, fire hazards, and
the like. Thus, refineries are usually located to take
advantage of the lower cost of crude-oil transportation, and
designed to produce a product slate that matches a local or
regional demand mix.
The same principles apply to petroleum (naphtha and
gas-oil) -based petrochemicals manufacturing. Crude oil is
cheaper to transport than the primary and intermediate
petrochemicals or end-user products made from it. In
addition, the initial distillation of crude oil and the subse-
quent cracking or reforming of naphtha or gas oil produce a
great variety of hydrocarbons. Some of these products are
suitable for petrochemical use, but others are more valuable
as gasoline, jet-fuel, or fuel-oil blending stocks. Thus,
petroleum-based petrochemical plants are generally planned
Economics and Alaska Outlook Page 115

as a part of refinery complexes, or are at least located near

refineries. As a result:
Transportation economics do NOT favor
Alaska locations for petroleum refineries (except
to serve in-state demand) or oil-based petro-
chemical plants.
These principles help explain why oil-industry person-
nel and energy analysts from the beginning almost unani-
mously doubted the economic viability and financeability of
the Alpetco proposal, both in its original petrochemical-
plant incarnation and in its later refinery version.
Natural gas and the lighter natural-gas "liquids" like
ethane tend to be more costly to ship than their liquid or
solid derivatives. Generally, therefore, it makes sense to
convert methane and ethane to non-gaseous substances
before shipping them long distances.
Accordingly, gas-based methanol plants and ethane-to-
ethylene plants are almost invariably located in gas produc-
ing areas. As ethylene is itself a light gas, which can be
moved by sea only as a chilled liquid in costly cryogenic
tankers, it is usually processed further into liquid or solid
petrochemical derivatives such as ethylene oxide or poly-
ethylene before being transported to distant markets. As a
result:
If Alaska natural gas and ethane are to be
converted to petrochemicals anywhere, trans-
portation economics favor an Alaska plant loca-
tion.
This principle is the rationale behind the Dow-Shell
group's strategy. Methane or ethane would have to be ship-
ped by pipeline at relatively high unit costs, or by cryogenic
tankers at even higher costs, to feed petrochemical plants in
the Lower 1/8 or, say, Japan. Converting methane to
methanol in Alaska, or ethane to ethylene and then to
polyethylene, for example, would facilitate transportation
and hence reduce the final cost of the chemical products.
Page 116 Economics and Alaska Outlook

Final-products manufacturing. The advantage of loca-

ting gas-based hydrocarbon-processing facilities near feed-
stock sources does not extend indefinitely "downstream".
Just as in other Alaska resource-based industries - - - wood
products and fisheries, for example --- the state's compara-
tive advantage in manufacturing ends with those kinds of
processing that reduce shipping costs by decreasing the bulk,
weight, or perishability of the product. For a long time to
come, the more complicated, labor-intensive, and weight- or
bulk-increasing manufacturing activities will be cheapest to
carry out in populous areas close to major markets. For this
reason:
Alaska petrochemicals manufacturing will
probably end with first or second derivatives
that can be shipped as liquids or solids for
further processing and fabrication elsewhere.
High capital, labor, and transportation costs make it
unlikely, in other words, that a petrochemical complex in
Alaska would produce and package fibers, textiles, apparel,
housewares, rubber or rubber products, pharmaceuticals,
etc. Although the public-relations literature of the various
chemical companies emphasizes the vast number of final
products made from, say, ethylene derivatives, the Dow-
Shell reports make it clear that the group is not actively
considering processing Alaska hydrocarbons beyond the first
form in which they can be shipped economically to other
markets.

Fixed Capital Costs.

The foregoing principles in themselves do not guaran-
tee that it is economically feasible to make petrochemicals
in Alaska from North Slope hydrocarbon gases. Fixed
capital costs --- essentially plant construction costs --- are
also a crucial element in investment and plant-location
decisions. It is conceivable, therefore, that the the con-
struction-cost disadvantages of an Alaska location might
overwhelm its transportation-cost advantage. For the same
reason, relative transportation expense alone does not
Economics and Alaska Outlook Page 117

necessarily dictate where in Alaska a plant should be

located.
Because refining and petrochemicals are unusually
capital-intensive industries, however, production labor and
other operating costs are relatively unimportant.
Fixed costs vs. variable costs. No new hydrocar-
bons-processing facility is likely to be built unless its spon-
sors and their lenders are convinced that project sales
revenues will be sufficient to cover the full cost of produc-
tion; that is, to recoup both (a) fixed costs --- the entire
original investment plus a competitive return on that invest-
ment --- and (b) variable costs --- feedstock costs and
other operating expenses.
Because an individual plant or complex costs hundreds
of millions or even billions of dollars, cost overruns and
mistaken product-market or feedstock-supply forecasts can
be catastrophic, so that investors normally demand that
their feasibility studies demonstrate a substantial safety
margin. Therefore:
Investors in a NEW plant will insist that
expected sales revenues cover fixed as well as
operating costs,
Nevertheless, once a plant is built, sunk
costs do not affect operating decisions.
Lumped together, these two propositions may seem
confusing, but they are relatively simple concepts. An
established plant will tend to operate at virtually full
capacity so long as its product sells for more than its
feedstock and other operating costs, even if it is not
covering its depreciation or debt service, or generating any
net profit.
In such a situation, in other words, the goal of plant
management is to minimize losses rather than to maximize
profits; refineries or chemical plants will stay in service
whenever they would lose more money by shutting down
than by continuing to operate. Such plants would never
Page 118 Economics and Alaska Outlook

have been built, however, if their sponsors had expected

them to operate at a chronic loss.
The outlook for refinery investments. A huge over-
hang of excess refining capacity now exists both in the U.S.
and worldwide. As a result, the current pr ices of petroleum
products tend to represent little more than the cost of
feedstocks, and contain no allowance for the amortization
of fixed costs or any return to even the capital invested in
existing facilities. Such a market is Jess likely yet to
provide the substantial margin above operating costs that is
necessary to justify building new refineries, unless they have
some exceptional offsetting advantage in the form of low-
cost feedstocks, captive markets, or a direct government
subsidy.
Oil refining is thus a money-losing business almost
everywhere, and it is likely to remain a money-losing
business for many years. This situation probably would have
been fatal to the Alpetco refinery scheme even if it did not
have to face Alaska's transportation and construction-cost
disadvantages.
The outlook for petrochemicals investments. New in-
vestments in facilities to produce ethylene and ethylene
derivatives might seem to face the same difficulties as
refinery investments, because great excess plant capacity
now exists for olefins both nationally and worldwide. The
crucial difference in outlook between refining and petro-
chemicals, however, is that there is little prospect that the
expansion of oil-product consumption will resume in the
foreseeable future, but most analysts believe that petro-
chemical consumption will begin growing again when the
present economic slump ends. There is also a reasonable
chance that the prices of gaseous feedstocks in remote
producing regions like Alaska or Saudi Arabia will be suffi-
ciently lower than world prices for competing oil-based
feedstocks to make new gas-based plants profitable even in
the face of idle capacity elsewhere.
The "Alaska cost differential". Big construction pro-
jects in Alaska are notorious for their high costs relative to
Economics and Alaska Outlook Page 119

their counterparts in more developed temperate regions.

Industrial facilities in Alaska must be designed to withstand
more severe environmental stresses. At the same time,
labor expense and transportation charges for equipment and
materials are higher in Alaska, while labor productivity is
lower than in the Lower 48, Europe, or East Asia.
Local construction expenses, chiefly site preparation
and on-site labor costs, are usually assumed to be 50 to 60
percent higher at tidewater in Southcentral Alaska (e.g., at
Anchorage, Kenai, or Valdez) than on the U.S. Gulf Coast;
about JOO percent higher in Interior Alaska (Fairbanks or Big
Delta); and about three times as high in the Arctic (Prudhoe
Bay or Barrow). The Dow-Shell group has concluded that
the cost differentials for petrochemical-plant construction
in Alaska are even greater than these, projected plant costs
in Southcentral Alaska ranging from 1.7 to 2.1 percent of
the Gulf Coast cost. (These figures, incidentally, are
comparable to the typical cost differential for refinery or
chemical-plant construction in the Middle East.) Therefore:

If a processing plant in Alaska is to be

competitive, the sum of its transportation and
feedstock-cost advantages must be sufficient to
overcome a large construction-cost disadvan-
tage.

Feedstock Costs and Feedstock Supply.

The costs of feedstock and fuel (which are often but
not always the same) are a crucial factor in deciding the
feasibility of any refinery or petrochemical investment.
Oil-based feedstock costs. Low ocean-transport costs
have created a single world market for crude oil, in which
prices everywhere move more or less in unison, and in which
differences in the price of crude oil between various tide-
water locations around the world are relatively small.
Petroleum refining and petroleum-liquids-based petro-
chemicals manufacturing therefore tend to be "price-taker"
Page 120 Economics and Alaska Outlook

industries. Long-term crude-oil or petroleum-product sales

contracts at fixed prices, or even at fixed formula prices
(say, at the Saudi Arabian "marker-crude" price plus or
minus a location and quality differential) are very rare.
Individual operators of petroleum-liquids processing plants
thus have little opportunity to control their raw-materials
costs, but typically must accept whatever prices world
markets (or government regulators) dictate. This is the case
even for a refiner or chemical producer that owns and
processes its own crude-oil supplies, because the true index
of feedstock costs to such a producer is the price the oil
might have commanded on the open market.
In assessing the economic feasibility of a fuels refin-
ery or oil-based petrochemical plant, therefore, its sponsors
have to make judgments about future oil prices and their
relation to the market value of the fuels or petrochemicals
derived from them. This task is not quite as hopeless as the
turbulent history of world oil prices might suggest, because
the market prices of petroleum-derived products from com-
peting plants will also vary with the price of crude oil. And
although substantial volumes of petrochemicals are pro-
duced from feedstocks other than crude-oil fractions, petro-
chemicals derived from oil constitute the "marginal" supply
--- the strategic portion of world output whose costs will
determine the product-price levels at which any new chemi-
cal plant must be competitive.
The feasibility analysis for a new hydrocarbon-liquids
processing plant need not concentrate on the absolute level
of oil prices, therefore, but only on:
The cost of feedstocks for the proposed
plant, RELATIVE to the expected costs for its
competitors (e.g., the difference between naph-
tha prices in region A and gas-oil prices in region
B).
The effect of oil-price levels on total pro-
duct demand.
Economics and Alaska Outlook Page 121

The Alpetco project, for example, would clearly have

lacked any advantage under the first test whether its
product was to be petroleum fuels or petrochemicals. Un-
less Alaska were willing to sell royalty oil at less than
market value, project sponsors had no reason to expect their
oil feedstock costs to be decisively lower than those of
Lower-48 or East Asian refiners or oil-based petrochemical
manufacturers.
As a refinery, at least, Alpetco was handicapped under
the second test too: Higher oil prices were persuading
consumers worldwide to reduce oil consumption, idling a
high proportion of existing refinery capacity in the United
States, the Caribbean, Europe, and East Asia. The result
has been --- and will continue to be --- petroleum-product
price levels that reflect near-zero operating profits for
refineries everywhere. Unless the State sold its crude-oil at
a very deep discount, therefore, no hope would exist for
Alpetco to recover its investment or earn any return on it.
Gas-based petrochemical feedstocks. Natural gas and
natural-gas liquids markets are quite different from those
for crude oil. Because of high costs for marine transport of
liquefied gases, a single world market for methane or ethane
does not exist as it does for crude oil and petroleum
products, while the market for LPG's is far less developed
than the crude-oil market. Even within North America, the
huge investments necessary to bring Arctic gas to market
would foster large regional differences in the wellhead value
of natural gas and gas liquids.
Accordingly, gas-based petrochemical plants in remote
producing areas like Alaska or the Middle East are likely to
be "price-makers" rather than price-takers. This means,
simply, that local petrochemical manufacturing may be able
to offer gas producers a higher price than they would get by
shipping the gas to distant markets by pipeline or as LNG.

Illustrations.
Natural gas is interchangeable with fuel oil in most of
its end uses. Because more than half of the natural gas
Page 122 Economics and Alaska Outlook

currently sold in the United States is consumed by electric

utilities and industry, the ultimate market value of addition-
al gas to Lower-48 consumers is the price of the residual
fuel oil that gas would displace.
Let us ignore for the moment the complications crea-
ted by federal wellhead price controls and "rolled-in" pric-
ing, and suppose that the value of North Slope natural gas
used as fuel in the Lower 48 is roughly equal to the price of
residual oil at $30.00 per barrel or $5.00 per million btu
(mmbtu). If the cost of transporting North Slope gas to
Lower lt8 consumers is $4.00 per mmbtu, its netback value
on the North Slope would only be $1.00 per mmbtu. Thus,
the gas producers would gain if they could obtain any price
above $1.00 for the North Slope hydrocarbons that would be
worth $5.00 in the Lower 48. Figure 9 illustrates the way in
which transportation costs determine the netback value of
natural gas.

Figure 9
Netback Value of Prudhoe Bay Natural Gas

Natural-Gas
Lower-48 Markets Pipeline Prudhoe Bay

Natural-gas value
($5.00 per mmbtu) less pipeline
Netback gas value
equals the transportation
($1.00 per mmbtu)
oil price cost ($4.00
($30.00 per barrel) per mmbtu)

Methanol and MTBE at Prudhoe Bay. In 1980, a group

that included the Arctic Slope Regional Corporation
(ASRC), Davy-McKee International (DMI), and Westinghouse
proposed to study the feasibility of building an industrial
complex at Prudhoe Bay that would produce methanol
and MTBE (methyl tertiary butyl ether), a high-octane syn-
thetic gasoline for shipment either through TAPS or
through a new light-liquids pipeline.
Economics and Alaska Outlook Page 123

The viability of a North Slope methanol/MTBE concept

would depend on whether (1) the ability to obtain feedstocks
for a small fraction of the Lower 1/8 price ($1.00+ vs. $5.00
in the preceding illustration) would offset (2) the higher cost
of transporting the petrochemical products to market plus
(3) the higher capital cost of construction in the Arctic.
(Because of the highly-automated, capital-intensive charac-
ter of the plants, higher operating costs would not be a
major factor.)
NGL-derived olefins in Southcentral Alaska. Under
the same assumptions as in the previous examples, the gas
producers would have three options for selling the ethane
from their North Slope natural gas liquids:
(l) The ethane could be sold to gas-trans-
mission companies for shipment to the Lower 1/8
as part of the pipeline-gas stream at a wellhead
price of $1.00;
(2) The ethane could be extracted from the
pipeline-gas stream for use as plant fuel at
Prudhoe Bay, thus freeing additional methane
that is worth $1.00 when shipped through the gas
pipeline, or
(3) The ethane could be extracted from the
natural gas either on the North Slope or at
Fair banks for shipment through a new pipeline to
a petrochemical complex at Valdez, Anchorage,
or Kenai.
Suppose further that the cost of moving the ethane
from the North Slope to the petrochemical complex were
$2.00 per mmbtu. If ethane were worth any more than $3.00
as a feedstock for petrochemicals manufacturing, a petro-
chemical-plant operator could afford to offer the North
Slope gas producers a better price for their ethane than they
would receive under either of their other two options.
Page 124- Economics and Alaska Outlook

This, basically, is the Dow-Shell concept. The eco-

nomic viability of the proposed NGL's pipeline and olefins
plant would depend on whether the plant's ability to get
feedstocks at a cost significantly lower than the Lower-4-8
price would be sufficient to offset the capital-cost disad-
vantage of operating in Alaska and any transportation-cost
disadvantage for the petrochemicals produced.
Feedstock value vs. feedstock price. The value of
North Slope methane or ethane as pipeline gas (or conceiv-
ably, LNG) in distant markets typically establishes its
opportunity cost in any sale for use as a petrochemical
feedstock in Alaska, but this cost does not automatically
determine the market price.
Figure 10 illustrates this situation. ( Note: The
figures in this illustration are wholly hypothetical, and not
purported to be realistic.) Suppose as before, that a gas
producer ("Exxon") could get $LOO per mmbtu for ethane
shipped to the Midwest via the natural gas pipeline, and that
the same ethane would cost $2.00 to ship from Prudhoe Bay
to an ethylene plant in the Cook Inlet region. And suppose
further that a chemical company ("Dow") determines that it
could produce petrochemicals worth $8.00 ($9.00 in the
Lower 4-8, less $1.00 transportation cost) in that plant from
each mmbtu of ethane feedstock at a manufacturing cost
(including a normal profit) of $4-.50 Thus the value of the
ethane as petrochemical feedstock at Cook Inlet would be
$3.50 per mmbtu ($8.00 less $4-.50), and "Dow" would make
an acceptable profit so long as it could get the feedstock for
$3.50 or less.
"Exxon's" opportunity cost of $1.00 per mmbtu plus
transportation costs of $2.00 would, therefore, establish a
$3.00 floor price at the ethylene plant, while the $8.00
netback value of the petrochemicals less "Dow's" $4-.50
manufacturing cost would establish a $3.50 ceiling price.
Figure 10 asks at what price between the floor and the
ceiling would "Exxon" be likely to sell its ethane to "Dow";
specifically, how would the extra 50 cents per mmbtu in
potential profits be shared between them.
In the specific case at hand, there is reason to expect
the actual sales price to be nearer to the ceiling (ethane's
Economics and Alaska Outlook Page 125

Figure 10
Value vs. Opportunity Cost at Cook Inlet

Lower-48 Markets Prudhoe Bay

Natural-gas value Ethane opportunit

($5.00 per mmbtu) less gas-pipe- cost
equals the line trans ort equals the
oil price cost ( 4.00 netback gas value
($30.00 /barrel) per mmbtu) ($1.00 er mmbtu)

Cook Inlet

plus NGL's pipe- Ethane

line trans ort opportunity cost
cost 2.00 to "Exxon"
per mmbtu) ($3.00 er mmbtu)

What is the ethane

Oil- or gas-based price? (Who gets
feedstocks the extra 50 cents
($6.00 er mmbtu) per mmbtu profit?)
($3.50 less $3.00)
plus manufacturing
cost ($3.00
t
Value to "Dow":
per mmbtu) ethane value as
feedstock
petrochemical ($3.50 er mmbtu)
market value
($9.00 er mmbtu)
+
less petrochemical
manufacturing
cost ($4.50
per mmbtu)

netback value of
petrochemicals
per mmbtu) ($8.00 er mmbtu)
Page 126 Economics and Alaska Outlook

value to "Dow") than to the floor ("Exxon's" opportunity

cost). The reason is that "Exxon" itself could become an
Alaska petrochemical producer and reap the full profit if
"Dow's" highest offer were less than the value of the ethane
as a petrochemical feedstock.
The real-life Exxon is itself one of the world's biggest
petrochemical producers, as are British Petroleum and Arco.
If "Dow" determined that the most-likely value of North
Slope ethane as feedstock for a Cook Inlet ethylene plant
were $3.50 per mmbtu, it is reasonable to suppose that the
chemical subsidiary of "Exxon" or some other North Slope
gas producer would also conclude that ethane was worth
about $3.50 as a feedstock for a Cook Inlet ethylene plant.
It is not likely, therefore, that "Exxon" would "leave money
on the table" by selling feedstocks to a competitor for less
than this price.
Alternative scenarios. It is instructive to examine two
variations on this scenario: (I) in which the Alaska gas
pipeline is not built, and (2) in which wellhead price controls
on Prudhoe Bay natural gas influence the producers' deci-
sion on the disposition of North Slope hydrocarbons.
(I) No gas pipeline. It is conceivable that North
Slope natural gas will cost more to ship to the Lower 1/8 by
pipeline than it would be worth as fuel when it arrived, so
that its netback value at Prudhoe Bay would be close to zero
or even negative. Or, alternatively, the gas pipeline may
not be built for some financial or political reason. In either
of these cases, it would not be the value of ethane as part of
a gas stream destined for the Lower 1/8 that would establish
the producers' opportunity cost for North Slope ethane. The
absence of the gas-pipeline alternative would establish a
floor price for ethane feedstocks at approximately the cost
of extracting it from the produced natural gas, which would
otherwise be reinjected or (if the law permitted) flared.
If "Exxon's" added out-of-pocket cost of extracting
and gathering the ethane were 75 cents per mmbtu, the
company's floor price for ethane at Cook Inlet would be
$2.75 ($0.75 plus $2.00 transport cost). But so long as
Economics and Alaska Outlook Page 127

"Exxon" itself had the option of processing the ethane into

ethylene, the probable sales price of ethane would not be
affected by "Exxon's" loss of the option to ship it out in the
gas pipeline; that sales price would still reflect the $3.50
value of the ethane as petrochemical feedstock in Cook
Inlet, just as it did in Figure 10.
This comparison illustrates our earlier remark that a
gas-based petrochemical plant could be a raw-materials
price-maker rather than a price-taker like an oil-based
petrochemical plant. So long as ethane's feedstock value is
higher than its opportunity cost or its out-of-pocket cost to
the producers, that value will tend to determine the price.
(2) The effect of wellhead price controls: meth-
ane. The Natural Gas Policy Act of 1978 establishes a
ceiling price for Prudhoe Bay natural gas. This price is a
little bit higher than $2.00 per mmbtu in 1981 and is
scheduled to escalate with general inflation.
Interestingly, the federal ceiling price may be con-
siderably higher than the netback value of North Slope gas
at the wellhead (its market value in the Lower 48 less
processing, transportation, and distribution costs). Without
the regulation of Lower-48 natural-gas prices, the North
Slope producers probably could not hope to collect the
ceiling price for their gas. The gas producers are neverthe-
less counting on price controls for Lower-48 natural gas plus
"rolled-in" pricing to make North Slope gas marketable at
the federal ceiling pr ice.
Let us assume for the sake of illustration that these
hopes are realistic, and that Alaska gas could be sold in the
Lower 48 on a rolled-in basis at a price of $8.00 per mmbtu
instead of the $5.00 we assumed in Figures 9 and JO. This
instance is illustrated in Figure 11. The netback value of
North Slope gas is now $4.00 per mmbtu ($8.00 in the Lower
48 less $4.00 transportation expense), in contrast to $1.00 in
the earlier examples. If the wellhead price is subject to a
federal ceiling of $2.00, however, the ceiling price (rather
than the $4.00 netback value) would establish "Exxon's"
opportunity cost for the gas.
Page 128 Economic s and Alaska Outlook

Figure 11
Methane-to-Methanol Price Determination

Lower- 48 Markets Prudhoe Bay

Final- market gas Netback gas value

value ($5 .00
pe r mmbtu)
(not a licable) less gas-pipe-
I
_ _ _ _ _____,_..,~ ($1.00 per mmbtu)
(not a licable)

line transport Netback gas value

Rolled-in gas cost ($4.00 ($4.00 per mmbtu)
value ($8.00 per mmbtu) _.. exceeds ceiling
per mmbtu) (not a lic able)
(not a licable )

I Ceiling price
($2.00 e r mmbtu)

plus me thanol
manufacturing
cost ($2.50)
f
plus pipe line North Slope meth-
tr a nsporta ti on anol opportunity
cost (= $1.00 cost to "Exxon"
pe r mmbtu (= $4.50 per mm-
of me thane) btu of me thane )

Cook Inlet
W'

Lower- 48 market Cook Inlet meth- Extra profit to

value of methanol anol cost ..._either "Exxon" or
( = $7 .00 per mm- (= $5.00 per mm- -"Dow" is 50 cents:
btu of methane) btu of methane) "Exxon" is not
indifferent between
less t a nker trans- Netback meth- selling to "Dow"
__Q_ortation (= $1.00-"" anol value and sellin__g_ t o
per mmbtu of ($6.00 per mm- its own affiliate
methane btu of me tha ne
Economics and Alaska Outlook Page 129

Let us further assume that each mmbtu of methane

could be converted on the North Slope into methanol at a
manufacturing cost (including a normal profit) of $2.50, and
shipped to Cook Inlet at a cost of $1.00, where the methanol
would be worth $6.00. Thus, the netback value of North
Slope methane would be $4.00 ($8.00 less $4.00) if it were
shipped to the Lower 48 as pipeline gas, and only $2.50
($6.00 less $2.50 manufacturing cost and $1.00 pipeline
transportation cost) if shipped as methanol. "Exxon", how-
ever, would be forbidden to charge more than $2.00 for the
feedstock in either case. How would the gas actually be
used, and what would its actual sales price be?
Interestingly, the existence of a ceiling price would
tend to favor using the gas as raw material for methanol
production. Because the use could not itself affect the
wellhead price, "Exxon" as producer of the gas would be
indifferent to whether it was sold for shipment through
ANGTS or for processing in Alaska. Because the gas is
worth $2.50 as a chemical feedstock, however, any chemical
company to whom Exxon sold its gas at $2.00 would reap a
50-cent windfall. The obvious course for "Exxon" would be
to avoid arm's-length sales entirely, and to sell its North
Slope gas to its own chemical subsidiary.
(3) The effect of wellhead price controls: ethane. If
ethane and other NGL's were extracted from Prudhoe Bay
natural gas in the field before they entered an interstate
natural gas pipeline, they would not be subject to federal
ceiling prices under the Natural Gas Policy Act or any other
provisions of federal law governing "natural gas". However,
any gas liquids shipped through ANGTS (and thereby "com-
mingled" with methane) would lose their separate legal
identity and become subject to the same price controls as
the methane. Indeed, though the question is legally argu-
able, the Federal Energy Regulatory Commission (FERC)
and the federal courts may deem ethane commingled with
the sales-gas stream to be "commodities in interstate com-
merce" and, accordingly, subject them to even further
federal regulation --- even if they are ultimately separated
Page 130 Economics and Alaska Outlook

from the pipeline gas and sold in Alaska for processing

within the state.
The extent of federal regulatory jurisdiction over
North Slope ethane can not be forecast precisely today, but
it is possible that the Federal Energy Regulatory Commis-
sion (FERC) and the Department of Energy will try to claim
some jurisdiction over their prices, transportation, and/or
end-uses. Moreover, there will surely be some private
parties and perhaps State agencies in the Lower 1/8 demand-
ing that the Commission exert such jurisdiction.
All in all, it appears that shipment of NGL's through
the gas pipeline would subject them to wellhead price
controls and possibly to direct federal controls over their
transportation logistics and ultimate disposition. If North
Slope ethane shipped through ANGTS were extracted, say,
at Big Delta for shipment through an NGL's pipeline to a
petrochemical plant in Valdez or Kenai, federal regulation
could create the same kind of incentive for the producers to
avoid an arms-length sale (to Dow-Shell, for example) but
rather to maintain control of the feedstock, just as we
described with respect to price-controlled methane.
The foregoing scenarios have been for the purpose of
illustration only. They hardly begin to exhaust the roster of
economic and regulatory possibilities. The pricing assump-
tions in these examples, moreover, were not intended to be
realistic (though the following section will explore some of
the issues in understanding what constitutes realistic fig-
ures). And finally, the Prudhoe Bay gas producers may have
considerably less control over the disposition of their natu-
ral gas than some of the examples suggest, because they
have already sold that gas, at least conditionally, to Lower-
1/8 gas pipeline companies. Nevertheless, these scenarios
point to facts about the disposition of North Slope oil, gas,
and NGL's that are often overlooked in Alaska:
(1) It is not clear that ANY system for
marketing or processing North Slope gas and gas
liquids --- be it ANGTS, LNG, the Dow-Shell
Economics and Alaska Outlook Page 131

concept, or methanol --- will be organizational-

ly, legally, and economically viable in the fore-
seeable future.
(2) The ultimate disposition of the differ-
ent North Slope hydrocarbons and their alloca-
tion among pipeline sales gas, petrochemical
feedstocks, and field fuels will be be determined
mainly by how the gas producers perceive their
own interests, and how they view the viability of
the various systems that might be proposed.
The Northwest Alaskan partnership, the State of Alas-
ka, potential outside purchasers like Dow and Shell, or the
ventures of Alaska Native corporations can have com-
paratively little influence on either of these circumstances.

Economies of Scale
"Economies of scale" are another factor that combines
with transportation economics, feedstock prices, and rela-
tive construction costs to influence owner decisions where
to locate various kinds of industrial facilities. The term
refers to the tendency of larger machines, plants, firms, and
industries to have lower average unit costs of production,
processing, or transportation than smaller units of the same
kind. Several elementary physical principles contribute to
technical economies of scale in the equipment and plants
used for petroleum refining, pipeline transportation, and
petrochemicals manufacturing. Three of most important of
these principles are as follows:
The amount of steel in a pipe increases
roughly in proportion to its diameter; but the
volume of fluid it can contain increases with its
cross-sectional area (which is proportional to the
square of its diameter) and, because the amount
of friction depends on the inner surface area
(which is directly proportional to the diameter
rather than to the cross-sectional area) the flu-
ids-carrying capacity of the pipe increases more
than proportionally to the square of its diameter.
Page 132 Economics and Alaska Outlook

The amount of steel in a refinery or chemi-

cal-plant processing vessel, and its heat-loss (or
gain) by radiation, are proportional to its surface
area (which increases with the square of its
length or width), while the volume of fluids it
can hold during processing is proportional to the
cube of any one of its dimensions.
Increasing the size of a given piece of
equipment does not necessarily require any in-
crease (and almost never requires a proportional
increase) in its operating and supervisory man-
power, or in the investment in control-system
equipment.
A common (if imprecise) rule of thumb with respect to
both process equipment and pipelines is that fixed costs tend
to increase with the six-tenths power of capacity. That is,
if a 50 mb/ d refinery costs $500 million, a comparable 100
mb/d refinery can be expected to cost about $760 million
($50 million x 26). Thus, doubling the refinery's size
reduces its fixed cost per unit of capacity 24 percent, from
$10,000 per barrel per day, to $7 ,600.
Comparable rules of thumb are (a) that operating labor
requirements vary by the one-fifth power of capacity, and
that fuel consumption varies with its four-fifths power. If a
50 mb/d refinery needs 100 workers, therefore, a 100-mb/d
refinery would need about 115 workers (I 00 x 22); a
doubling of capacity would increase total fuel requirements
by about 76 percent (28 = J.76)), meaning that average fuel
costs per barrel would fall about 12 percent (1.76/2 = .88).
Limits to economies of scale, and the optimum scale.
Economies of scale always have upper limits dictated by
physical, economic, or human factors. The size of refinery
or chemical-plant process vessels, for example, is limited by
the strength of materials, safety considerations, and the
disruption that would be caused by the planned or unplanned
shutdown of a single large unit rather than one of a series of
smaller units.
Economics and Alaska Outlook Page 133

Thus, there tends to be some optimum (or lowest unit-

cost) size for each kind of facility. Actually, the optimum
size in most cases tends to be a rather broad range of sizes,
over which unit costs at a given rate of capacity utilization
(say, 90 percent) are rather flat. There always seems to be
a region, in other words, where the economies of scale in
some parts of the system just about offset diseconomies in
others.
The optimum scale for complex fuels refineries nowa-
days is in the 100-to-250 mb/d range. The optimum scale
for oil tankers in intercontinental service (e.g., between the
Persian Gulf and the U.S.) is 250 to 500 thousand tons (mt),
but the optimum size on shorter hauls (e.g., Valdez to Puget
Sound) is considerably less. The reason is that larger
tankers take longer to load and unload. As a result, they
would spend an uneconomic proportion of their total ser-
vice-lives in port if they were used on short hauls, thereby
dissipating the operating economies of scale that they
acheive only while sailing.
The optimum size for ethane crackers is between 1
and 1.5 billion pounds per year, and the optimum size for oil
and gas pipelines seems to be in the 42-to-56-inch range.
Most of these figures tend to increase over time, as a result
of the development of stronger steels.
"Worldscale" facilities are facilities built to the tech-
nically optimum size because they have not had to be scaled
to a limited feedstock supply or product market. They are
facilities, in other words, that have the entire world to
draw on and/or to serve and which, because they are of the
optimum size, are able to compete in world markets.
Collier Carbon and Chemical Company's ammonia/
urea plant at Kenai is worldscale; Alpetco proposed a world-
scale refinery at Valdez; and the Dow-Sheil study is con-
sidering a worldscale petrochemicals complex. Alaska's
existing refineries, however, are not worldscale: their
design scale was keyed to the limited size of the Alaska
market rather than to national or international competition.
Page 134 Economics and Alaska Outlook

Analyzing Project Feasibility

Refinery and petrochemical plant investment decisions
depend on several variables, including the ones that this
chapter has considered in some detail (transportation, con-
struction, and feedstock costs, and economies of scale), plus
a number of others that include:
Feedstock requirements
Feedstock characteristics
Process engineering and
operation
Fuel and energy supplies
and prices
Labor, materials, utilities,
and services needs and
prices
Product slates and volumes
Product pr ices
Capital structure
Interest rates
Inflation rates
Federal, state, and local
taxes
Health, safety, and environ-
mental regulation
A companion volume to this report, Zinder Energy
Processing, "Preliminary Economic Evaluation of NGL-
Based Petrochemical Production in Alaska", October 1980,
provides a useful accounting framework for most of these
variables, and a rudimentary economic model for relating
them to one another with respect to a project similar to
that contemplated by the Dow-Shell group.
Economics and Alaska Outlook Page 135

The final product of most economic feasibility studies

includes (I) a "pro-for ma" income statement, and (2) a dis-
counted cash-flow (DCF) or "internal" rate-of-return analy-
sis. A pro-forma income statement lists and sums
hypothetical figures for the major cost and revenue
elements for each year over the project's economic life,
usually 20 to 25 years. A DCF analysis calculates the rate
of return on investment (ROI) implied by the whole stream
of negative and positive cash-flow figures in the income
statement. Judgments on project feasibility, then, depend
upon the DCF rate-of-return estimate: Is the expected
return high enough to justify the investment?
Economic feasibility reports vary greatly in sophisti-
cation and detail, depending on the project and sponsor on
the purpose of the report. In general, a preliminary
"reconaissance" study will use far more general assumptions
and simpler models than a report prepared for prospective
lenders, who usually insist on a completed engineering
design and detailed market analysis, among other things.
Assessing uncertainty and risk. Greater methodologi-
cal sophistication and detail do not necessarily improve the
quality of an economic feasibility study, however. The most
critical factors determining the economic feasibility for
refineries and petrochemical plants are often judgmental
assumptions, for which the most rigorous engineering or
econometric methods give little hope of precision. Worse
yet, the very complexity and appearance of precision in a
formal economic analysis tend to obscure the intuitive and
imprecise insights about the key unknowns that are
absolutely crucial to intelligent decision-making.
The most powerful variable in determining the outlook
for new worldscale hydrocarbons-processing plants is the
outlook for world oil prices. By influencing petroleum-
product and petrochemical prices, oil prices determine the
level of consumption and the rate of demand growth. Since
the prices of crude-oil, natural gas, LPG's, coal, and other
feedstocks do not necessarily move together, the outlook for
crude-oil prices is central, not only to
Page 136 Economics and Alaska Outlook

deciding whether any new plants should be built, but to

choosing the right combination of raw-materials and plant
location.
Expert opinions about how crude-oil prices will change
over a period as short as five years vary by a factor of two
or even three. Certainly, the recent history of petroleum
prices calls for some humility in forecasting the future. The
weighted-average wellhead price of Prudhoe Bay crude oil
has quadrupled since mid-1978, reaching an all-time peak of
about $26 per barrel In March 1981; it has now (August 1981)
fallen below $22, and few analysts would be startled by 1985
prices as low as $15 or as high as $50. Because of the
crucial role oil prices play in determining both the costs and
revenues of any new refinery or petrochemical plant, the
range of uncertainty about oil prices probably overwhelms
the influence of all other economic assumptions combined.
The level of oil prices and the many other variables
that are powerfully influenced by oil prices are not the only
factors that are essential inputs to any feasibility analysis
yet subject to horrible uncertainties. Capital-cost
estimates for large construction projects are notoriously
unreliable --- TAPS would have been the largest economic
debacle in U.S. history if its huge cost overruns had not be
offset by an even larger, unanticipated, leap in the price of
imported oil. There is, likewise, no scientific way to
determine future inflation rates.
Unfortunately, the feasibility reports of major energy
projects that are offered to investors, government officials,
and the public, are usually designed primarily as means of
persuasion rather than business-decision tools. Most such
reports present a single pro-forma income table and DCF-
return-on-equity figure, concluding that the "most-likely"
return-on-equity (ROE) from the project in question is, say,
15 percent. At most, the authors will offer several
"scenarios" that correspond to different prepackaged sets of
assumptions and show different ROE's, but which give the
reader Ii ttle basis for choosing one scenario over another.
Economics and Alaska Outlook Page 137

Sensitivity analyses. A relatively simple device

for improving the usefulness of feasibility analyses, but one
which has been absent in the public literature regarding
Alpetco, the Alaska Highway gas pipeline, and the like, is a
sensitivity analysis that tells the reader which assumptions
are truly critical and how critical they are. Even the Zinder
report regarding NGL-based petrochemicals in Alaska, cited
above, fails to tell its readers how its final results would be
affected by a $5.00 per barrel change, plus or minus, in
world oil prices; by a given percentage construction-cost
overrun; or by a specified change in the project's capital
structure, or in interest rates, etc. The Zinder model does
allow users in state government to vary inputs one by one
and to observe any change in the result, but this computing
capability is no substitute for a clearly presented sensitivity
analysis.
Risk analysis. One step in sophistication beyond
sensitivity analysis is risk analysis, which explicitly incorpo-
rates uncertainty into its calculations. If experts are willing
to attach probability figures to their assumptions, a "Monte
Carlo" or "decision-tree" risk-analysis program will produce
conclusions in terms of probabilities. A risk analysis can
offer the investor or public official a much more powerful
decision tool than a single "most likely" figure, or even a set
of "high", "medium", and "low" estimates.
The risk analysis of a given hypothetical project might
begin with a set of expert judgments on the probability
distributions of cost overruns, completion delays, future oil-
price trends, product-market conditions, interest rates, and
general inflation, and conclude as follows:
"There is a 50-percent probability that the
DCF ROE will be 15 percent or higher." (This
means, of course, there is an even chance that it
will be lower than 15 percent.) "There is, how-
ever, a 20-percent risk that the ROE (return-on-
equity) will be zero or less, and a IO-percent risk
that the project will default on its debt."
Equity investors might consider a 15-percent profit
expectation (the weighted average of all probable outcomes)
Page 138 Economics and Alaska Outlook

as adequate, and be willing to accept a one-in-five risk of

losing money if this risk is offset by a "fair gamble" of
earning much more than 15 percent. The JO-percent risk of
default would probably be intolerable to prospective lenders,
however. The risk analysis might also say of the proposed
investment, that:

"Changing the debt-equity ratio from 7 5:25

to 50:50 would reduce the chance of default
from one-in-ten to one-in-fifty; the probability
of just breaking even or losing money would fall
from one-in-five to one-in-ten. But reducing the
"leverage" (the proportion of debt) in the pro-
ject's capital structure in this way would also
reduce the expected ROE from 15 percent to
11.5 percent."
In this case the risk of default might be low enough,
but the expected ROE would be inadequate. Combining risk
analysis and sensitivity analysis gives us an even more
powerful decision tool. Consider this observation about
another fictional project:

"Although the expected rate of return and

risk of default are both acceptable, we must
point out that this project will never make a
profit in the unlikely event that world oil prices
stablize at their current levels or continue to
decline. To achieve a 50-percent expectation of
a 15-percent ROE, we must assume that oil
prices advance at an average annual rate at least
two percent faster than general inflation."
The intuition of the investor or policymaker on how
"unlikely" it is that world oil prices will stabilize may be
just as good as that of the experts who carried out the
analysis. In any event, the user of the analysis now has the
information with which to make his own policy judgment.
Finally, risk analysis could offer the following kind of
observation on a hypothetical royalty-oil sales proposal:
Economics and Alaska Outlook Page 139

"The proposed project has a better-than-

even chance of standing on its own feet. In
order to reduce the probability of default to less
than 5 percent so that private debt financing can
be obtained, however, the State must be prepa-
red to discount its royalty oil by as much as
$5.00 per barrel if and when necessary to meet
debt-service demands. The likelihood that a
subsidy of this magnitude will be necessary is
less than 7 percent but there is an almost one-
third chance that some discount on feedstocks
will ultimately be required."
The contract between the State of Alaska and the
Battelle Northwest Laboratories analyzing the proposed
Susitna hydroelectric project and its alternatives requires
Battelle to provide a full range of sensitivity analyses, to be
specific about the probabilities assigned to key assumptions,
and to present its results in the form of probability distribu-
tions. To our knowledge, this is the first time the State has
made such an assigment --- but the Susitna project is one
that could involve a direct outlay of billions of dollars of
State money.

Coping with uncertainty and risk.

The analytical methods described in the previous pages
do not reduce or control business risks, but only identify and
attach numbers to them. Means do exist, however, by which
refiners and petrochemical manufacturers can reduce their
exposure to surprises and the damage caused by them. The
chief measures are long-term contracts, plant and system
flexibility, vertical integration, horizontal concentration,
and risk-spreading through diversification.
Long-term contracts. Investors in refineries and
petrochemical plants can reduce their capital costs and
certain kinds of business risks by building a highly specializ-
ed facility designed to process a single feedstock into a
predetermined product slate for a predetermined customer
or group of customers.
Page 1110 Economics and Alaska Outlook

This kind of arrangement is much more common in oil

and gas transportation and among utility companies than it
is in either refining or chemicals manufacturing. A prospec-
tive shipper on a proposed pipeline may offer the carrier
(the pipeline company) a "throughput and deficiency agree-
ment," under which the shipper promises to pay the carrier a
minimum bill proportional to the desired transport capacity,
even if the shipper does not use that capacity.
Likewise, a utility that buys coal, natural gas, or
electricity may bind itself in a "take-or-pay" contract to
pay for a specified volume on a specified schedule, whether
or not the utility actually takes the contracted amount. A
stricter version of the minimum-bill or take-or-pay contract
greatly facilitates debt financing. This is the "all events" or
''hell-or-high-water" provision, which requires the shipper or
purchaser to pay the minimum bill even if the carrier or
seller can not perform (because of project noncompletion or
breakdown, for example).
Facilities with long-term "back-to-back" raw-
materials purchase and product-sales contracts are
generally easy to finance with very high debt-to-equity
ratios. One example is the Alberta Gas Ethylene Company's
Joffre plant, which has long-term contracts from its parent
(Nova, Ltd.) for ethane feedstock, and a long term "cost-of-
service" ethylene sales contract with Dow Canada.
Project financing. One advantage of projects with
back-to-back purchase and sales contracts is that they can,
at least in principle, be "project-financed" with "nonre-
course" debt. Project financing establishes a new corporate
entity to own and operate the facility and the non-recourse
feature means that the facility's owners are not responsible
for debt service; their exposure is limited to their equity
contribution, which may be comparatively small. (Capital
structures for most project-financing proposals tend to
contain 7 5 to 90 percent debt. The original financing plan
for the Alpetco plant contemplated no net equity contribu-
tion by its sponsors.)
Economics and Alaska Outlook Page 141

Project financing is not a method, however, of elimi-

nating risk but only of shifting it to other parties through
take-or-pay or similar contracts, and its feasibility depends
both upon the creditworthiness of those parties and the
tightness of their contractual obligations. For this reason,
it is mostly regulated public utilities that use this financing
technique, and the technique is feasible even for them only
where State and Federal regulatory agencies can assure in
advance that debt-service charges will be "perfectly
tracked" to a captive market of final consumers.
The sponsors of many, if not most, recent large-scale
energy-industry ventures have hoped that they could pro-
ject-finance them with a high ratio of nonrecourse debt ---
the Alaska Highway natural gas pipeline, Alpetco, and the
Northern Tier oil pipeline are familiar examples. Very few
have ever been successful, and we are not aware of any
completely nonrecourse financing that has yet been carried
out for a major nonutility energy project. Successful
financing has always seemed to hinge on a creditworthy
third party that agrees to pay off the debt even if the
facility is never completed.
The success ratio of attempted project financings over
the last few years has been so low that promoters of huge
and complex energy projects are increasingly seeking, in-
stead, government loan and purchase guarantees or joint
ventures with large equity participants. Nevertheless, two
maxims will be useful to Alaskans in evaluating future
industrial promotions:
(1) Lending institutions are not willing to
bear the completion, technical, and marketabil-
ity risks for large-scale resource-extraction,
transportation, or processing ventures in Alaska;
and
(2) Unless the sponsors have found
someone able and willing to provide the project
with conventional levels of equity capital (25
percent or more) and to guarantee the project's
entire debt (at least until it goes into operation),
it is reasonable to assume that the facility will
not be built.
Page llJ.2 Economics and Alaska Outlook

In other words, nonrecourse and highly-leveraged pro-

ject financing is as much a myth for enterprises in Alaska as
it has proved to be in the Lower IJ.8. Even the participation
of major international companies (like U.S. Steel,
Westinghouse, and Getty Oil in the Northern Tier Pipeline,
or Exxon, Arco, and Columbia Gas in ANGTS) does not
assure financing unless these companies are willing to supply
risk capital and guarantee project debt. However, certain
features of project financing, like long-term contracts that
facilitate plant specialization can provide a real boost to
more conventional strategies of capital formation. The rest
of this chapter covers some other techniques for reducing
business risk.
Plant and system flexibility. Although plant special-
ization tends to reduce technical risks and construction
costs, it magnifies feedstock-supply and market risks. Most
refineries and petrochemical complexes built in recent years
have considerable built-in flexibility. In the case of refiner-
ies, technical flexibility provides the ability to run a wide
range of crude-oil mixtures and to vary their product slates.
Petrochemical complexes have been built, where feasible, to
include (or provide room for future addition of) both an
ethane cracker and a naphtha cracker.
A large company with several plants of different
design, adapted to different feedstocks and different pro-
duct slates, will have much more flexibility to deal with
changes in raw-material supply and market conditions than a
smaller, single-plant enterprise, even if the large firm's
individual plants are relatively specialized.
Horizontal concentration. It is the advantages of
system-wide flexibility that encourage horizontal concen-
tration --- the tendency of big firms to get bigger. In the
l980's, for example, Dow and its affiliated ventures and
joint ventures will be producing (or buying on long-term
contract) ethylene from naphtha and gas oil on the U.S.
Gulf, in Europe, and in East Asia, and ethylene from NGL's
produced and marketed under a number of different ar-
rangements in the southwestern United States, Alberta,
Saudi Arabia, and perhaps Alaska.
Economics and Alaska Outlook Page 143

Obviously, not all of these ethylene supplies will turn

out to be the lowest-cost sources, and some of the ventures
may well prove to be money-losers. But with a broad,
diversified feedstock base, Dow is unlikely to do worse than
the average in its industry, and unless the world market for
ethylene derivatives stagnates completely, Dow should do
very well in the next decade.
Vertical integration. An earlier chapter of this report
alluded to the historical tendency of crude-oil producers to
integrate downstream into refining and products marketing
in order to assure themselves product outlets and thus to
retain their market shares in periods of surplus. BP's
acquisition of Sohio and part of the Sinclair system is a
doubly outstanding example - - - first because of the obvious
logic of the combination and second because it was only
partly successful.
With the discovery of the Prudhoe Bay field in 1968,
BP was about to become North America's number one crude-
oil producer, but it had no refineries or retail outlets and
hence no assured market. Sohio, on the other hand, was the
largest "independent" refiner --- independent in the sense
of having almost no crude-oil production. To this extent the
BP/Sohio marriage was perfect.
The geography of the merger has turned out to be
abominable, however, particularly in light of Congressional
restrictions on foreign exchanges of North Slope crude oil.
Unlike Arco and Exxon, which have refineries and dealer
networks on the West Coast, Sohio has none, and thus the BP
group still has no properly situated outlets for its crude oil.
As a result, Sohio has to absorb two or three dollars per
barrel in added transport costs for oil sold or exchanged east
of the Rockies --- a burden that Arco and Exxon are spared
on most of their Alaska production.
"Upstream" or "backward" integration of refiners or
petrochemical companies into crude-oil production not only
gives the processor a more secure raw-materials supply, but
helps stabilize feedstock costs as well. During the first half
of 1981, for example, the most recent round of OPEC price
increases together with the deregulation of domestic crude-
Page 144 Economics and Alaska Outlook

oil prices raised the average cost of raw materials to U.S.

refiners several dollars per barrel, but market conditions did
not allow them to recover these higher costs in petroleum-
product prices. This situation put most independent refiners
and refiners with low crude-oil self-sufficiency ratios into a
no-profit or operating-loss position. To the extent that a
refiner is self-sufficient in crude oil, however, the loss of
refining profits is at least partially offset by additional
crude-oil production profits. (The offset is usually less than
total, because of the higher royalty, severance-tax, and
Windfall Profits Tax liabilities that result from higher
wellhead prices.)
The unreliability of foreign crude-oil supplies in recent
years has made upstream vertical integration highly attrac-
tive at the same time that it has become increasingly
expensive to achieve. The decline in Lower-48 crude-oil
production and the major companies' loss of overseas con-
cessions have drastically reduced self-sufficiency ratios.
After several Middle-Eastern oil-supply interruptions, al-
most every large refiner and petrochemical producer, re-
gardless of its existing degree of backward integration, has
been trying to get direct control of as much crude oil as its
can or, failing that, to work out some kind of marriage or
joint venture with a crude-oil producer. With the acquisi-
tion of Sohio by BP in the early 1970's, Ashland and Clark
are the last large independent refiners.
In 1977-78, for example, Ashland attempted to sell a
major share of the company to the National Iranian Oil
Company (NIOC) in exchange for a long-term crude-oil
guarantee. In 1979 Getty Oil --- despite its surplus of
crude-oil production relative to its refinery capacity ---
bought Reserve Oil Company, whose subsidiary Western
Crude Oil gathers and markets crude oil for hundreds of
small producers. Just this year, the Hawaiian Independent
Oil Company announced a major investment by Kuwaiti
interests, who would presumably be responsible for providing
crude oil to the Hawaii refinery.
The Alaska NGL-based petrochemical scenario set out
in connection with Figure 10 earlier in this chapter offers a
Economics and Alaska Outlook Page 145
final illustration of the risk-reducing value of vertical
integration. In this scenario, a prospective feedstock seller
("Exxon") and a prospective buyer ("Dow") are both confi-
dent that the value of Prudhoe Bay ethane as feedstock for
an ethylene cracker at Cook Inlet is about 50 cents per
mmbtu more than its value as part of the sales-gas stream
in the Alaska Highway natural-gas pipeline.
Neither party really knows what the NGL's extraction
plant and pipeline or the petrochemicals plant will cost, or
what world market conditions will be for ethylene deriv-
atives five or ten years from now. Accordingly, the
anticipated profit per mmbtu of ethane shipped and pro-
cessed, while assuredly positive, might be considerably less
than 50 cents or considerably more than 50 cents. But any
feedstock price low enough to insure "Dow" against loss
would be considerably less than the "most-likely" value of
the material according to "Exxon's" estimate. The obvious
resolution of this dilemma would be for Exxon to sell its
NGL's not to "Dow" but to an "Exxon" subsidiary (or perhaps
to a "Dow-Exxon" joint venture). This way, "Exxon" would
receive the whole profit (or nearly the whole profit), whe-
ther it turned out to be large or small.
Another incentive favoring vertical integration comes
from uncertainty about transportation charges on the NGL's
pipeline. It is impossible to know for certain to what extent
the Federal Energy Regulatory Commission (FERC) and the
Alaska Public Utilities Commission (APUC) will regulate
charges on the pipeline or what rules they will use. Federal
regulation of oil pipelines has historically used a "fair-value"
rate base, which permits charges to increase over time, but
a FERC administrative law judge recommended in 1980 that
TAPS transportation tariffs be set on the basis of "depreci-
ated original cost," which results in declining charges. The
choice between the two rules may vary the first-year
transportation charges on an Alaska NGL's line by a factor
of two or three. Uncertainty about pipeline charges thus
may lend great uncertainty to any assessment of the feasi-
bility of a petrochemicals industry based on NGL's in
Alaska.
Page 146 Economics and Alaska Outlook

Once more, resolution of the dilemma may hinge on

vertical integration. If the major shipper owns the pipeline,
the tariff as such does not much matter (apart from its
effect on tax and royalty collections) --- pipeline trans-
portation charges are largely a bookkeeping shift of profits
(or losses) from one pocket to another. The most risk-
protected system for Alaska petrochemicals, therefore,
would include producer participation in both then NGL's
ownership and in the petrochemical complex.