Você está na página 1de 11

May 1959

Petroleum R Plant and


Equipment Ex 1950-58

VEN in this age of rocketry the finding, not necessarily less interesting or important,
E drilling, and processing of oil are ven-
tures which capture the imagination. One
aspects of the oil industry have captured the
interest of economists and students of busi-
need not hold an oil lease to quicken at the ness in general. Those interested in the pric-
sight of a several-hundred-foot-high gusher. ing process, particularly the relation between
Nor need one be overly methodical to stand cost and price, are intrigued by the problems
in awe of the quiet efficiency with which the involved in manufacturing a basic raw mate-
modern refinery daily processes barrels of oil rial-crude oil-into literally a thousand fin-
by the hundreds of thousands. Finally, one ished and semi-finished products. Similarly,
need not be a gambler to be impressed with students of the tax structure find an equally
the risk involved when upwards of a million intriguing set of issues concerned with the in-
dollars is ventured in drilling a wildcat well centive effects of depletion allowances on do-
which may not even gurgle, much less gush. mestic and foreign production, or the equity
In addition to these more dramatic facets considerations which such tax treatment
62 of the industry, certain less well-known, but ra1ses.

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

May 1959 MONTHLY REVIEW

But it is perhaps in the field of investment place economically obsolescent facilities. In


analysis that the oil industry presents some of practice it is often extremely difficult to make
its most interesting and provocative prob- this distinction because rarely does the flow of
lems. This. is particularly true of petroleum capital spending fall exclusively into either
refining. In this industry technological devel- category. A somewhat more detailed classifi-
opment has been so rapid and so consistent cation would distinguish between new facili-
that it gives the appearance of a continuing ties that are virtually identical with older cap-
innovational boom. This persistent techno- ital stock, and equipment that either provides
logical advance has in tum left its mark on the the same output at different real costs or pro-
annual plant and equipment expenditures for vides qualitatively different products from
refineries. The impact of innovation has been those produced by older facilities. Again, this
such as to eliminate any notion of a rigidly apparently clear-cut conceptual distinction is
defined relationship between purchases of not easily confirmed empirically.
plant and equipment and the level or change When the new and old plant and equip-
in refinery output. ment are virtually identical, in terms of qual-
Since both capital-saving and capital-using ity of product and real costs of production at
innovations are strong and persistent ele- all levels of output, then we say that the new
ments in the refining industry's technological facilities do not change the production func-
pattern, the capital-to-capacity ratio may be tion. When, on the other hand, new plant and
viewed as subject to continuing, offsetting, equipment changes quality, cost, or product
pres~ures -the capital-saving innovations mix, we speak of an innovation. In the ab-
constantly working to lower the ratio, the sence of innovation we should expect fixed
capital-using innovations working to raise the capital outlays to manifest a regular, but not
ratio. Although the long-run trend has been necessarily changeless, relation to the level of
an increasing ratio, year-to-year changes output and capacity. We would not expect
have frequently moved against this trend. The this relationship to be rigidly fixed, of course,
direction and magnitude of these changes because businessmen inevitably will over- or
have produced an investment pattern which underestimate their capital requirements, or
seriously challenges the hypothesis that "the" may choose for short periods to respond to
capital-to-capacity ratio tends toward some increased demand by running plants at
equilibrium value. Moreover, the apparently greater-than-optimum operating rates. But
volatile and unpredictable nature of these despite these inevitable irregularities, the cap-
technological changes suggests that the causes ital-output and capital-capacity relationships
are grounded neither in basically cyclical nor should tend toward some normal value.
secular forces, but are significantly deter- In contrast, an industry possessing a dy-
mined by such industry characteristics as the namic technology is capable of changing mar-
structure of the refining industry, the kind and ginal capital coefficients by 50 to 100 percent
quantity of competition, the nature of the within a few years. Under these circum-
commodities produced, and the inherent in- stances, an explanation and analysis of the
novational possibilities. investment pattern requires an understand-
ing, first, of the technological change which
Plant and Equipment Outlays- conditions the investment process, i.e., inven-
s· . tions and their application; and, second, of
Plant and equipment expenditures may be the conditions which have encouraged or in-
distinguished, theoretically, in terms of duced the technological change. Let us look
whether they add new capacity or simply re- at these conditions in the oil industry. 63

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

FEDERAL RESERVE BANK OF SAN FRANCISCO

In the petroleum refining industry there is technological advance has become an increas-
little doubt that competition, while intense ingly important factor in this decline.
and often bitter, has nevertheless increas-
ingly been manifested in various nonprice Petroleum Chemistry and Refining
forms such as product and process innova- •
tion. These innovations have taken the form For our purposes the essential technologi-
of changes in the refiner's product mix, im- cal consideration is that oil refining means
provements in product quality, and reduced converting and transforming a thick black
processing costs. The pace of such activity liquid (crude oil) into an indefinite number of
has increased substantially in recent years. products, the most important of which are
The petroleum industry employed 17,000 gasoline, distillate, and residual fuel oils.
scientific and engineering personnel in 1956, This transformation is extremely expensive
and was spending over $130 million annually and, though suggestive of magical processes,
on research and development programs. By is actually the commercial translation of
1958 research and development expenditures well-known principles of organic and physical
had climbed to $259 million, and planned ex- chemistry. Crude oils are essentially mix-
penditures for 1959 total $295 million. 1 tures of carbon and hydrogen compounds.
The impact of these programs has been Although the carbon content of almost all
striking. Over the last forty years technolog- crudes is close to 85 percent by weight, and
ical change has rendered basic refining equip- hydrogen constitutes almost all of the re-
ment obsolete on the average of once every mainder, no two crudes (from different
five years, 2 and so fundamental have been the fields) are identical. Part of the difference is
corresponding changes in yields and quality due to the varying oxygen, nitrogen, and sul-
that noninnovating refiners have been forced fur content, amounting to from less than one-
either to imitate their pioneering rivals or get half to almost five percent by weight. But the
out of the industry. basic reason is the variation in the molecular
Firms have been increasingly forced to size, kinds, and numbers of carbon-hydrogen
match the innovational pace set by the most compounds found in different crudes.
progressive, with some survivals of this race These differences broadly determine the
tending to become specialty plants serving amount of gasoline, kerosene, asphalt, etc.,
which are yielded by distillation. Fifty years
particular product needs or geographic areas.
ago the molecular mixture was the ultimate
For example, small asphalt plants are "shel-
determinant of the refiner's product mix.
tered" against the processing economies of
How much kerosene, fuel oil, or gasoline was
large plants by the heavy transportation costs produced was a simple function of the kind of
which a large plant must meet in selling to a crude being processed. Today these differ-
geographically widespread market. Similarly, ences are largely overcome by special process-
small fuels plants may find their "niche" by ing equipment, and it will be helpful to con-
selling to nonbrand gasoline jobbers, or by sider, however briefly, the principles and
selling nonpremium quality fuels to major re- techniques of petroleum refining.
finers for further processing. Nevertheless, the The basic refining process is distillation.
number of small refineries bas declined rap- Before World War I this simple process often
idly. Apparently the rapid and costly pace of constituted the entire refining operation. The
1 United States Bureau of Mines, Mineral Facts and Problems, coming of the automobile, and more recently
Bulletin 556 (1956), p . 671 , and McGraw-Hill Department of
Economics (see Business Week, April 18, 1959, p. 29). of high-compression engines, transformed
• W. L. Nelson, "Cracking Equipment Obsolescence," Oil and Gas this "basic" technique into preparation for
64 Journal, (September 9, 1957), pp. 123-4.

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
Federal Reserve Bank of St. Louis
http://fraser.stlouisfed.org/
Digitized for FRASER May 1959

HEAV IE R-THAN-GASOLI Nil!:


CoMPOUNDS
~
0

-
FROM OTHER PROCESSES
DISTILLATES FROM
"<

CRUDE DISTILLATION CRACKING


-o
VI
OR 11 11 -o
CRACKS, ALKYLATION
VACUUM DISTILLATION How
COMBINES LIGHT H1GH OcTANE G....sOLINII:

HYDROCARBON MOLE- I& MADI!.


TO
VACUUM CULES TO FORM VERY &IMPLif"IE:D ,-LOW CHART
DISTILLATION
.HIGH OCTANE GASo- 0,. A MODERN RC,.INII:RV
INCREASES VOLUME LINE COMPOUND CALL-
UES.
OP' MATERIAL P'ROM

CRUDEDISTILLATIO

THAT MAY BE P'ltD TO HIGH OCTANII GASOLINII

CATALYTIC CRACKER BLENDING (94 OCTANE)

~
0
z
-t
HuJH OCTANE GASOLINE :I:
r-
CRUDE DISTILLATION
(85-1 00 OCTANE) -<
STRAIGHT-RUN QAIH)LINIE
"'<m
m
~
~-T~f'lG~j,.._~l...:_"'HIGH OCTANE GA&OLINIE
(92 OCTANE)

CRuDE I"ROM WEL.Ls

-................
'"''""'·······.......
OCTANE GASOLINE

(96 OCTANE)

•,•.(V.-.'1''-VA

HIGH OCTANE GASOLINE


NUMBERS.
(80 -110 OCTANE)

0.

""
May 1959

FEDERAL RESERVE BANK OF SAN FRANCISCO

the more complex cracking and reforming the sales potential of "high" octane gaso-
processes. Distillation provides the refiner line and have spurred the addition of proc-
with relatively fixed amounts of gasoline, essing capacity which is primarily for prod-
light and heavy oils, and a sprinkling of light, uct - particularly octane - improvement.
even gaseous, products. Historically the Such processes include, in the jargon of the
changing market for gasoline, distillate, and trade: thermal reforming, catalytic reform-
residual fuels has been reflected in the devel- ing, tetraethyl lead addition, desulphuriza-
opment of refining processes to rearrange the tion, and isomerization. Additionally, such
molecular composition of products in con- processes as polymerization, alkylation, and
formance with market requirements. Where hydrotreating may be employed to raise the
once the products of the distillation columns octane rating of the gasoline produced. Alky-
were piped directly to market, the modem re- lation, for example, actually converts non-
finery subjects these products to an .increas- gasoline-type compounds into a gasoline of
ingly complex sequence of supplemental very high octane rating. Because it converts
processing. Some of the supplemental (in re- nongasoline molecules into gasoline-type
finery jargon, "downstream") processes con- compounds, it is, strictly speaking, a conver-
vert petroleum gases into automobile or avi- sion rather than improvement-type process.
ation fuel. Others transform thick, almost But if the high octane gasoline so produced
pitchy, distillate residue into diesel fuels , is a necessary blending stock for the manu-
heating oils, coke, and even gasoline. On the facture of high octane motor or aviation gaso-
other hand, many of the more recent proc- line, the process is in fact being employed as
essing developments have been geared not so an improvement device. Similarly, hydro-
much to converting as to improving the qual- treating1 can be primarily employed to change
ity of refinery products. product yields, or to improve product qual-
Product improvements have not, of course, ity. An estimated 70 percent of hydrotreating
been confined to developing better gasoline. capacity is employed for product improve-
But gasoline is the refiner's prime product, ment purposes.
and most of his product-improving dollars
have gone into motor fuel processing. These
dollars have developed processes for chang- The introduction or increased utilization
ing the gravity, suHur content, vapor pres- of these product improvement processes gave
sure, distillation range, and octane rating of shape to the refining investment pattern of
gasoline. the early and mid-1950's. The pattern was
Although octane ratings are not the sole that of boom, and the dimensions of the boom
index of gasoline quality, they have com- are readily suggested by the Plant and Equip-
manded most of the refining and motoring ment Expenditure Series in Chart 1. Fixed
public's attention. This attention has not been 1 Hydrotreating is a generic term covering the use of hydrogen in
a wide variety of refining operations including sulfur and nitro-
undeserved. From the motorist's point of gen removal, increasing oxidation stability, color improvement,
etc.
view, octane improvement is one of the keys 2 The difficulties of giving empirical content to conceptual rela-
tionships are painfully evident in the following section. Al-
to increased engine performance potential. though various estimates of capital expenditures are available.
only one source provides estimates of plant and equipment ex-
From the standpoint of the refiner it has be- penditures closely conforming to the relationship which we can
use in this section. We need estimates of plant and equipment
come increasingly expensive as octanes in- expenditures that count only those outlays which result in o'/ler-
atmg capacity within the calendar year cited. Only the Depart-
crease, and consequently productive of a ment of Commerce's Annual Survey of Manufactures provides
data on this basis, and because refinery construction schedules
growing number of technologic and eco- often extend over two or three years we cannot use capital ex-
penditure for plant and equipment which does not result in oper-
nomic headaches. Petroleum marketing de- ating capacity for one, two, or more years. Unfortunately, the
Surveys of Manufactures covering 1957 and 1958 are not yet
66 partments, however, are acutely aware of available.

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

May 1959 MONTHLY REVIEW

CHART 1 other. That is, except for changes in equip-


PLANT AND EQUIPMENT EXPENDITURES ment costs or changes in the technique of pro-
OF PETROLEUM REFINERIES
MILLIONS OF DOLLARS 1960·156 duction, the expenditure series should par-
100 allel the series for additions to gross capacity.
·-···: :::· Instead, after adjusting for cost changes, the
soo
expenditure series increases during a year
when capacity added "declines" (1952) and
300 declines when gross capacity added increases
(1954) . In addition to the dissimilarities in
200
turning points, the rates of change in the two
series exhibit considerable differences. For
example, in 1955 capacity added declined
more sharply than did capital expenditure .
.. JOO Table 1 reveals the extent of year-to-year
1950 1951 1952 1953 1954 1955 1956
SEMILOGARITHMET IC CHART variations in the ratio of plant and equip-
Source : United States Department of Commerce, Annual Sur'IJe:y
oj f!anufactures, Constant Dollar Series computed using Nel- ment expenditures to gross additions to crude
sons (Orland Gas Journal) Construction Cost Index 1946 = (distillation) capacity, from 1950-56.
100. I

capital outlays rose consistently over a five- The sharp changes from year to year pose
or six-year period, and culminated in a interesting questions. For example, why is
spending rate in 19 56 that was 100 percent the ratio for 1955 so much higher than that
(deflated) or 170 percent (current dollar for any other year; and why do the ratios for
series) over that prevailing in 1950 at the 1950, 1951, and 1956 stand apart from those
start of the boom.t for other years? The former ratios range from
These series bear little resemblance to the $4?1 to $434 per barrel of basic capacity,
crude capacity series in Chart 2. Yet, except while the latter range from $540 to $775 per
for technological change, we should expect a barrel. The primary explanation appears to
close and continuing similarity, for the one be the change in composition of investment,
would be the monetary reflection of the CHART 2
REFINERY GROSS (CRUDE CHARGING)
1
Although !den tical figures for ~957 are not available, comparable CAPACITY ADDED AND RETIRED
data published by the respective ecconomic departments of the
~h~se Manhattan Bank and McGraw-Hill Publishing Company BARRELS PER OAT 19~0-~6
md1cate the boom actually culminated in 1957, rather than 1956.

TABLE 1
RATIO OF P EQUIPMENT
EXPENDITURES TO GROSS

Expenditures Capacity Added


(1946 dollars) (barrels per day) Ratio
[1) [2] UH-[21
1950 174 million 401 thousand 434
1951 216 539 401
1952 274 451 607
1953 368 651 565
1954 364 673 541
1955 290 374 775
1956 352 818 430
100
Average 536 1950 1951 1952 1953 1954 1955 1956
SEMILOGARlTHMETIC CHART
Source : United of Commerce, Survey of Man·
~tates Depar.tment
ujactures, v~nous annual1ssues,
and United States Department Sour~e: United S~ates Department of the Interior, Bureau of
of the Intenor, Bureau of Mines, Information Circulars 7613 Mm es I nformation Circulars 7613 7667 7693 7724 7761
7646, 7667, 7693, 7724, 7761, and 7815. ' and781S. ' ' ' ' ' 67
Digitized for FRASER
http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

FEDERAL RESERV E BANK OF SAN FRANCISCO

as between outlays for basic and supple- indeed, it has become basic to his economic
mental capacity. Between 1953 and 1956 survival.
basic crude capacity increased about 11 per- In the 1920's and 1930's competitive po-
cent, while supplemental processing capacity sition was maintained primarily by introduc-
increased almost 50 percent. In 1953 there ing thermal cracking and thermal reforming
was less than one barrel of supplemental ca- equipment. During the 1940's the conversion
pacity for each barrel of distillation capacity. to catalytically cracked gasolines began, and
By 19 56 the ratio of supplemental to basic this conversion process continued into the
capacity bad increased almost 50 percent. 1950's. More recently the trend has been to
Chart 3 illustrates the changing relationship catalytically reformed gasolines. So swift has
between basic and supplemental processing been the adoption of new equipment that the
capacity. Apparently, then, the full explana- typical refinery of 1953 would be hopelessly
tion for the variable ratio lies in the tech- obsolete in the motor fuels market of 1959.
nological dynamics of a changing market. This is not the first time the refiner has
CHART 3
been faced with technological obsolescence.
PROCESS CAPACITIES- PETROLEUM REFINERIES Refining history has been marked by the de-
1950·58 velopment and improvement of such proc-
esses as continuous distillation, fractiona-
tion, cracking, reforming, polymerizing, isom-
erizing, and hydrotreating. Individually or
in combination these advances have under-
mined the economic position of the refiner
who failed to ride with the technological
tides.
In the 'twenties and 'thirties the introduc-
tion of thermal cracking and reforming pro-
500 foundly affected refinery yields and product
quality. Using the new cracking and reform-
ing equipment, the refiner of the early 1930's
200
doubled his gasoline yields, and at the same
POLY MERIUTIOII
IQO (OUTPUT C:APAc;ITTI ---........ time, produced a gasoline of increased octane
1950 1952 1954 1956 1958 rating.
SEMIL,OGARITHM!TIC C:HAitT
Again, in the 1940's the introduction of
Note : Stream days are days during which a pi:'OCes3 is actually In
full operation . Barrels per stream day represents maximum catalytic processes vastly altered the refining
daily average capacity ; It does not allow for necessary shut-
down time for routine main tena.nce, repairs, etc. art. During this period, however, little change
Source: Oil and Gas lrJUmal, Annual Refining Surveys, 1950-59.
was effected in gasoline yields. The primary
Technolog ical requirements for impact of catalytic cracking was to raise the
the cha ng ing ma rket refiner's "middle-of-the-barrel" yields-re-
We have referred to the impact of the motor ducing the relative yield of residual fuels and
age on refining technique. Where it was once increasing the yield of distillate fuel oils. Al-
an unmarketable byproduct, the automobile though the trend to catalytic equipment sig-
age has made gasoline the refiner's prime nificantly increased octane numbers, yield
product. No longer something to be surrep- rather than quality improvement was the
titiously dumped into nearby streams or primary basis for the trend. The technical su-
mixed with kerosene, gasoline bas come to periority of catalytic cracking stemmed pri-
68 determine the refiner's competitive position; marily from the fact that it enables the re-

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

May 1959 MONTHLY REVIEW

finer to reduce the relative yield of low-profit Plant A will be assumed to have no cata-
residual oil, increase the yield of middle-dis- lytic reforming unit, while Plants B and C are
tillates, and, at the same time, maintain yields provided·with 10,000 barrel and 27,000 bar-
of high quality gasoline. rel units. How do these differences affect gas-
Although octane improvement from 1920 oline yield and quality (octane number)?
to 1950 was not negligible (average octane Immediately apparent is the fact that the
increased from 50 to 84), and although at considerable octane-generating potential of
least a dozen octane improvement processes catalytic reforming involves a significant re-
antedate catalytic reforming, octane improve- duction in gasoline yield, all other things be-
ment was often primarily a valuable byprod- ing equal. But a brief glance at refinery yields
uct of technical change. The investment since the advent of catalytic reforming fails
boom of the 1950's is the first that can be to reveal this declining tendency. Indeed, gas-
characterized as reflecting expenditures de- oline yields have increased slightly. Since
signed mainly for motor fuels improvement catalytic reforming capacity has increased
-meaning, primarily, octane improvement. enormously it would appear that refiners have
Indeed, the introduction of increasing cata- offset the yield effects by introducing other
lytic reforming capacity, the mainstay of re- process capacity whose basic function is to
cent octane-improving techniques, has actu- maintain gasoline yields as quality is im-
ally tended to reduce gasoline yields. This is proved. The catalytic cracking, vacuum dis-
in marked contrast to earlier developments tillation, and hydrotreating capacity in-
when yield and quality improvements were volved in maintaining yield while octane
achieved simultaneously. number has risen with catalytic reforming is
The impact on yield and octane number is very costly and such expense has come to ac-
apparent if we compare the performance of count for an increasing share of the refiner's
three hypothetical refineries whose only dif- equipment dollar.
ferences lie in their respective reforming ca- The impact of these developments is per-
pacities. To keep our figures simple we shall haps best shown by comparing the process-
assume that the basic (crude-throughput) ca- ing capacities of the mythical "representa-
pacity of each refinery is 100,000 barrels per tive" refiner of 1954 and 1958. (Note that
average plant size grew significantly during
TABLE 2
these years, from 25,500 to 32,500 barrels
THE EFFECT 0 lNG CAPACITY per day.)
ON GASOLINE YIELD AND TABLE 3
CTA
PROCES ITI ES OF
Gasoline yield Plant A Plant B Plant C uREPRESENTATIVE" REFINERY,
(percent of total output,
including losses) 59o/. 56% 52 o/.
Average octane number (Barrels per day)
(including tetraethyl Percent
1954 1958 Increase
lead) 87 92 98
Crude capacity 25,500 32,500 27.5
Source : Oil and Gas Journal, (August 16, 1954), pp. 88·91. Vacuum distillation
capacity 3,700 9,800 164.9
Catalytic cracking 8,700 14,300 64.4
day, and that each possesses a catalytic crack- Catalytic reforming 765 5,200 579,7
ing unit of 48,000 barrel capacity. Each re- Thermal operations 7,752 7,345 -5.3
Hydrotreating 250e 4,600 1740.0
finery is assumed to add an identical amount Alkylation soo• 1,000 100.0
of tetraethyllead to its gasoline yield. (Addi- Polymerization 306 500 63.4
tionally these refineries possess other proc- •Estimated.
essing units which need not concern us.) Source : Computed from Oil and GIU Journal, annual refinery
surveys. 69
Digitized for FRASER
http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

FEDERAL RESERVE BANK OF SAN FRANCISCO

While the increase in average plant size is TABLB4

partly a reflection of expanding market de-


mand, and partly the result of fuller exploita-
tion of the tremendous economies of scale in- 10,000 30,000 60,000 100,000 200,000
Refinery size (Barrel s per day)
herent in petroleum processing plants, the
disproportionate increases in nonbasic (i.e., Process investment
per barrel 610 400 318 275 252
other than crude) capacity are primarily the Total investment
reflection of other forces. Although all non- per barrel 1 1,300 833 683 610 545
basic capacities increased proportionately 1
I ncludes: utilities, offsite facilities, tankage, and doc.ks.
Source: J. G. McLean and R. W. Haigh, Th e Growlh o/ bde·
more than the 30 percent increase in crude grated Oil Companies, (Boston: Harvard Unive,rsity, 1954),
pp. 558-567 .
capacity, it is noteworthy that the largest rela-
tive gains were registered by catalytic re- In 1954 the Oil and Gas Journal estimated
forming and hydro treating processes ( 5 80 to that approximately 7 5 percent of current re-
1, 7 40 percent). Product improvement thus finery investment was directed at quality im-
obviously affects refinery investment. provements and yield changes. In 1957 the
Unfortunately, it is not possible to deter- Petroleum Refiner estimated that $150 mil-
mine the precise dollar costs involved in lion of the estimated 1957 plant and equip-
bringing the average 1954 refinery up to ment expenditure was solely for catalytic re-
representative processing capacities for 1958. forming units. Another $30 million was esti-
As we have noted, a large proportion of the mated to be for alkylation units. Thus, these
new supplemental capacity was provided by processes alone account for almost a quarter
hydrotreating and catalytic reforming equip- of refinery investment. Finally, a leading au-
ment. Between 1954 and 1958 fully a dozen thority in the field of petroleum technology
new catalytic reforming processes were in- estimated that at the prevailing levels in 1958
troduced, and these introductions, coupled each one point increase in octane number in-
with the continuing design improvements in volved an additional industrywide investment
the initial units, significantly altered the capi- of $330 million.t
tal-to-capacity relationship. The degree of Despite the large number of variables which
change is suggested by the fact that between determine the capital-to-capacity ratio for in-
1949 and 1952 the investment requirement dividual plants, it is possible to make reason-
per barrel of new catalytic reforming capacity ably accurate overall estimates of this rela-
fell over 60 percent. Reinforcing the cost- tionship. On the assumption that a "mod-
reducing effects of design improvements, the ern" refinery is one possessing the same rela-
additions to reforming capacity have been in tive processing capacities as the "typical"
regions where scale economies are pro- 1958 refinery (above), and that capital re-
nounced (2000 to 5000 barrels capacity for quirements per unit of new capacity are as
reforming units, for example). An additional follows:
cost-reducing factor is found in the increase crude and auxiliary-$1 ,000/barrel
in scale of plant. Estimates provided by Mc- catalytic cracking-$ 300/barrel
Lean and Haigh suggest that the investment catalytic reforming-$280 /barrel
costs per barrel of capacity for a 200,000 hydrogen treating-$200/barrel
barrel per day plant are less than half as alkylation-$! ,200/barrel
large as those for a comparable 10,000 b arrel polymerization- $! ,000/barrel
refinery. 1 (Table 4) capital investment per barrel of capacity
1 J. G. McLean and R. W. Haigh, The Gr011Jth of I ntegrated Oil 1 Eugene Ayres, CotJrumer Bulletin , (Gulf Oil Corp., Sept. 1958) ,
70 Companies, ( Boston : Harvard University, 1954), Chapter 20. pp. 29-31.

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

May 1959 MONTHLY REVIEW

TABLE 5 esses. On the other hand, product improve-


ESTIMATED TION COSTS ments and yield changes have required the
OF MODERN REFINERY IN refiner to install increasing amounts of so-
SELECTED Y called supplemental processing equipment.
Estimated cost
per barre l of capacity
These additions have markedly offset the cost
Year (1950 costs) reductions accruing from design improve-
1910 $176 ments and increased scale economies. The in-
191 s 206
creasing ratio of capital-to-capacity is a meas-
1927 240
1935 350 ure of the combined effect of these opposing
1946 680 forces.
1957 920
Since these developments have not been re-
Source: 19 10-46 estimates compu ted from to tal cost data in Mc-
Lean and Haigh. 19 57 estimate comp uted on basis of Curry
flected by proportionate changes in refined
and Haag capital estimates and assu mption that typical refin-
ery possessed basic (crude charging) capacity of 32 ,500 barrels
product prices, and have apparently been ac-
per day, and the relative processing capacities as given in companied by a reduction in refinery profits
Table 3.
(percent return on net assets) the implica-
would be approximately $1,300. 1 Adjusting tion is that technological progress has been
to 1950 costs reduces this to $920/ barrel, passed on to the consumer in the form of
and enables us to compare this figure with the product improvement.
capital-to-capacity estimates provided by Gasoline is obviously only one of the re-
McLean and Haigh. (Table 5) finer's products and octane enhancement is
Although our definition of "modem" re- only one of the many product improvements
finery is not precisely the one employed by which have been made over the years. Yet
McLean and Haigh (their definition re- even if we limit our analysis to this single
quired that the refinery use the most modem index of product quality we can improve our
processes then available, while our definition
understanding of refinery investment sub-
requires only that the refinery be "average"
stantially. We can, for example, identify at
with regard to processing equipment), it is
clear that refining in the 1950's continues to least two supplemental processes which are
become more capital-using, i.e., the capital solely octane improvement processes. These
requirement per unit of capacity continues processes, catalytic reforming and isomeriza-
to increase. Thus the postwar refining ex- tion, account for at least 1.5 million barrels
perience has continued, and perhaps acceler- of process capacity added since 1950, and
ated, the trend established early in the history during particular years have amounted to up-
of refining. wards of 100 to 200 percent of the net change
in basic (crude charging) capacity. We
Conclusions should also recognize that a substantial pro-
Refinery investment in the postwar period portion of alkylation, hydrotreating, and
has been, largely, the reflection of two basic polymerization capacity installed is directly
and opposing forces. On the one hand, equip- or indirectly related to octane improvement.
ment designs have been significantly im- The installation of some vacuum distillation
proved, and process capacities markedly in- capacity is indirectly related to the yield prob-
creased. These changes have combined to re- lems caused by increasing the severity of re-
duce sharply capital requirements per unit of forming operations.
process capacity for most individual proc- Four million barrels is probably a con-
J Capital investment requirements estima ted by S. S. Curry and servative figure for the amount of supple-
D. 0 . Haag, ( Universal Oil Products Company), Petroleum
World and Oil, (October 24, 1957), p . 171. mental capacity added since 1950 solely or 71

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis
May 1959

FEDERAL RESERVE BANK OF SAN FRANCISCO

virtually for octane improvement. The sig- tive strength of capital-saving and capital-
nificance of this undoubtedly conservative es- using technological change.
timate is best understood by considering that In spite of difficulties in the analysis (due
basic crude capacity increased only 4 million to lack of engineering cost data) the evidence
barrels between 1950 and 1959. of Tables 1 and 5 is clear: the swift adoption
This would suggest that in the refining in- of new processes and of improved processes
dustry nonprice competition, in effect, carries for increased capacity suggests that invest-
over into the investment dimension, and that ment theory must include, at the minimum,
therefore such investment may be construed some expression of the relationships between
largely to reflect the competitive structure of capital coefficients, product quality, and ca-
the market. Since this investment is not re- pacity, and not assume an easy relationship
lated primarily to cyclical or long-run changes between capital investment and use of ca-
in output, it requires analysis in terms of rela- pacity .


I
ECOVERY in thecountry as a whole, which Housing starts in April held to the ad-
R had shown signs of slowing in January
and February, picked up momentum in March
vanced March level and, on the basis of first
quarter Federal Housing Administration ap-
and developed even greater forward thrust in praisals completed and contracts awarded,
April. High levels of manufacturing activity outlays for residential housing and public util-
lifted the Federal Reserve Board's index of ities construction will probably rise further.
industrial production two points for the third On the other hand, expenditures on highway
consecutive month, as inventory building in construction and industrial building have been
heavy goods industries pushed durable goods less encouraging. Retail sales reached a new
production to the boom level of early 1957 peak in March with a 1 percent gain in both
and brought increasing pressure to bear on durable and nondurable goods. April sales
the already high operating rates of the nation's held to the advanced March level (after ad-
steel mills. Output of large household appli- justment for seasonal factors) and were about
ances fell slightly in March but remained, 9 percent above a year earlier.
nevertheless, at near record levels, and build- Income and employment developments
ing materials industries continued to advance were generally favorable in March, and in-
as construction activity returned in March to creasingly so in April. Seasonally adjusted
its January high. Automobile assembly in- nonfarm employment increased 240,000 in
creased 100,000 units in April and rose in March and 3 70,000 in April. These, com-
early May to the highest weekly total in 1959. bined with successive lengthenings of the fac-
Nondurable goods production, although fail- tory workweek, elevated personal income 1
ing to advance in March, rose slightly in April percent in March, followed by a further 1
to a level about 6 percent above the prere- percent gain in April, to a record annual rate
cession high. Petroleum refineries, chemical of $372.7 billion. The latest employment
plants, and paper mills achieved record levels gains have put nonfarm employment within
of production; rubber mill output reached a 700,000 of the prerecession high and have
record in March but was held down in April brought steadily increasing pressure on the
72 because of strikes at major plants. unemployment rate. Though this statistic re-

Digitized for FRASER


http://fraser.stlouisfed.org/
Federal Reserve Bank of St. Louis