SPE

SPE 22015

Parametric Estimating of Oil and Gas Production Facilities Capital

Costs Worldwide
C.W. Karlik, ARCO IntI. Oil & Gas Co.
Copyright 1991, Society of Petroleum Engineers, Inc.

This paper was prepared for presentation at the SPE Hydrocarbon Economics and Evaluation Symposium held in Dallas, Texas, April 11-12, 1991.

This paper was selected for presentation by an SPE Program Committee following review of information contained in an abstract sUbmitted by the author(s). Contents of t~e paper,
as presented, have not been reviewed by the Society of Petroleum Engineers and are subject to correction by the author(s). The material, as presented, does not necessanly reflect
any position of the Society of Petroleum Engineers, its officers, or members. Papers presented at SPE meetings are subject to publication review by Editorial Committees of the Society
of Petroleum Engineers. Permission to copy is restricted to an abstract of not more than 300 words. Illustrations may not be copied. The abstract should contain conspicuous acknowledgment
of where and by whom the paper is presented. Write Publications Manager, SPE, P.O. Box 833836, Richardson, TX 75083-3836. Telex, 730989 SPEDAL.

ABSTRACf At the same time, awareness of when

certain parameters are not significant in
A methodology for a parametric approach determining capital costs is discussed.
to estimating facilities costs worldwide for Elaboration on how precision does not
new ventures is presented. Screening equal accuracy is presented; many
estimates then can be made with minimal parameters are needlessly considered in
data and in a timely manner. new venture economic assessments.
Development of a parametric approach The major results of this effort are the
consists of four phases: (1) The ability to estimate facilities costs, thus
identification of sources of data and the addressing multiple scenarios for future oil
gathering of that data; (2) Sorting of data; and gas development and the capacity to
(3) The mathematical analysis of that data efficiently screen estimates related to new
to quantify relationships that are intuitive; venture opportunities.
and (4) Reporting and updating of results
periodically.
INTRODUCTION
Observations include a review of how
"estimates of estimates" produce an Oil and gas projects have been developed
effective means of determining the future worldwide in widely diverse locations and
cost of new venture facilities. Parameters situations. The purpose of this paper is to
considered in providing estimates include, present a methodology for estimating the
but are not limited to, recoverable oil and gas production facilities capital
reserves, peak production rates, location costs with minimal effort and in a timely
(onshore or offshore), terrain, water manner. Order-of-magnitude estimates
depth, distance to infrastructure, etc. are products of this effort; therefore,
Expert systems/actual costs also generate economic screening of projects can be
databases for this methodology. done expeditiously and yet accurately.
References and figures at end of paper The resulting relationships will be general
45
 SPE 22015

but very adequate for economic evaluation water depth). While some relationships
of new ventures since oil and gas seem absurd, unique correlations might be
exploration parameters will vary the result. In the datasets which follow it
tremendously in comparison to facilities was thought that the oil and gas facilities
costs. costs possibly would correlate to one, or
more, of 109 parameters. Examples were:
A key concern encompasses the
competing efforts of attaining precision Water depth
and accuracy in doing estimates. Jacket weight
Unfortunately, especially when technical Design capacity
professionals do estimating, there is a Distance from sales point
tendency to describe the production Plant inlet pressure
facility to minute detail. A comfort level Storage tanker size
is attained in this process. However, Number of wells
accuracy has improved little in the process Reservoir size
since inputs remain vague. For example, Field life
to develop a prospect offshore West Area of site
Mrica a concerted effort toward Terrain
determining the fabrication site for a Country
platform, the equipment specifications, Elevation
etc. could be initiated during screening Etc.
economic evaluation. The quality and
composition of the produced oil or gas is Data is obtained through surveys of
a major unknown. Also, unless the overall industry sources, reviews of internal
field economics are very sensitive to project data and estimates. Table 1
capital costs, precision is accomplished contains a list of potential external data
unnecessarily. In most screening efforts sources. Table 2 depicts an example of a
which encompass early decision making, data collection form. Emphasis must be
lease bidding, farm-ins, determination of placed on gathering basic reservoir,
initial drilling and further delineation, production and facilities data with cost
order-of-magnitude accuracy only is information.
required. For purposes of this paper, an
order-of-magnitude estimate will be Examples contained herein encompass an
defined as having limits of approximately - initial data collection effort including 109
50% to + 100% with an 80% confidence separate parameters on each of 122
level. The resulting confidence limits are projects.
prediction intervals for cost values.
A major part of the effort in gathering
Reducing estimating resource usage will data is assurance that the data was for
be emphasized. At the same time similar parameters for each project,
retaining reality will be the norm. specifically in the cost information. Care
is taken to insure that the capital
IDENTIFICATION OF DATA expenditures reflect only facilities costs
SOURCES AND DATA GATHERING and not drilling costs, and is adjusted to
1990 dollars using specific industry cost
The first step toward regression analysis indices to remove the effects of inflation;
involves formulating an unbiased set of thereby a project adjusted cost is
data. One must determine the variables calculated.
and range of data potentially having
significant impact on capital costs. Data for company-operated projects are
collected from the economic evaluation
Initially one should consider parameters data and project reports and verified by
which are common to the types of respective project management teams. On
economic evaluations done in the past, but co-owner operated joint-venture projects,
it is important to introduce innovative project progress reports are used. Data
approaches (i.e., correlate offshore (as in from industry-wide published sources are
offshore platforms) project costs to used for projects operated by other
production design capacity regardless of companies. Most data prior to 1987 are
46
 seE 22015

excluded due to uncertainty of data variable, is a satisfactory approach to this

validity and uncertainty introduced for problem. This study reviewed multivariate
infl~tio~ factor}ng. When data on a single regression, but this does not show
proJe~t IS obtamed. from multiple sources, significantly improved results over single
assocIated data pomts (i.e., reserves and variable regression.
capital cost.s) are ~inked. Data likely to be
correlated IS obtamed from a single Linear relationships are explored but
source. usually do not produce substantive results.
Since many costs relationships are not
linear, regressions are performed on the
SORTING OF DATA natural logarithms of the data. The result
is a linear relationship between the
For a ~orldwide stu~ynumerous (>50) natural logarithms of the dependent and
coun~rIes should be mcluded for project independent variables. In these cases the
locatIOn.. The ~ata can be categorized by resulting equations take the following'
geographIc regIOns: Southeast Asia form:
Australasia, Europe, Africa and So~th
America. A compilation of preliminary In(y) = a + b*ln(x)
unadjusted data is made. ' or

The thoroughness of data collection is y = Axb

analyzed to determine a useful data set.
~ r.eview of the data. set for this paper where A=e a
mdlcates an approprIate reduction in the
data set to 98 projects, each potentially It is possible that other types of
h8:ving 12 parameters (Table 3). Upon relationships would also fit the observed
thIS data set subsequent analysis is data better, but the results are satisfactory
conducted. At this point review of data and continued research is unnecessary for
conte~t is manageable with the quantity of screening purposes; otherwise needless
data fIelds reduced. Next, preliminary time is spent determining estimated costs.
hypotheses are established, thereby setting
the ~t~ge for analyzing and verifying A by-product of determining the
SUSpICIOUS data. The test of preliminary coefficients of the equation is the
hypotheses is accomplished by correlating coefficient of determination (r 2). For
A

the data. the purposes of this study, a coefficient of

determination equal to 0.7 or greater is
The stage for statistical analysis of the deemed initially to be significant. The
data using single and multivariate greater the r 2, the greater is the
A

regression techniques is set. correlation. In order to measure whether

the fit was meaningful, a "t" test can be
run supporting the results and/or the
MATHEMATICAL ANALYSIS standard error multiples of predicted costs
reviewed.
Sta~dard regression analysis is used to
derIve an equation liking costs, Y, to The data was examined to determine the
various se.t(s) of variabl~s (parameters). relationships and degree of correlation
A curve fIt to the data IS made using the between specific parameters. The data
least squares method. The results are an and regression results are presented in
equation that approximates the d~ta and a Figures 1 through 5. Each of the graphs
measure of the "goodness of fit". show the relationship between a single
independent variable and the dependent
While multivariate regression examines variable.
the. relationship between the dependent
varIable (Y value) and a combination of Obvious relationships of Adjusted
i~dependent variables (X values). Simple
Facilities Capital Cost versus Total
lmear regression, which examines the Equivalent Reserves are used as an
relationship between the dependent example of how estimating curves can be
variable and a single, independent developed. Beginning with a worldwide
47
 SPE 22015

sample of data (98 projects), Adjusted Figure 6 depicts the timing and accuracy
Project Cost versus Reserves is analyzed expected. Depending on management
and depicted in Figure 1. A moderate philosophy a confidence level of 70 to 95
correlation existed for the worldwide data, percent could be applied to the range
but based on an accuracy range of no depicted at each phase of the process.
more than -50% to + 100%, the resulting This confidence level may also vary
regression line cannot be used for depending on the owner's experience for a
.estimating. By merely concentrating on a particular area.
single region, a satisfactory range can be
found. For example, Figure 2 showing Finally, a key to an effective estimating
Southeast Asian data produces a program is the periodic updating of the
satisfactory fit. data. New technology, governmental
policies and new resources are a few
Figures 3 through 5 depict an increasingly reasons for data obsolescence. Keeping
narrowing of criteria to produce another current is essential.
acceptable fit. The first pass (Figure 3)
covers European data; a segment of the Rather than narrowing in on specific types
worldwide data updated and augmented of projects and smaller geographic regions,
by further research also produces improved correlations may exist by
unsatisfactory results. Figure 4 contains establishing a list of assumptions thereby
the North Sea data for permanent restricting sample content. A typical
installations, better correlations appear series of restrictive assumptions are
but with too wide of an accuracy range (- summarized below:
60% + 160%). By only plotting fixed-
platform, United Kingdom data points a An offtake rate that provided a
relationship satisfactory for estimating is pseudo-life of 20 years for the field
produced (Figure 5). A curve for was assumed.
estimating is the result assuming the new No upstream costs have been included.
developments are within the scope of All plants are considered to be grass
recent activity (i.e., North Sea platforms in roots developments.
less than 200 meters of water). The latest proven technologies have
been included as the estimate basis
Also, consideration should be given to for each case.
adjustments for "economies of scale." All estimates will be provided as if
Note that the exponent of x is usually .s.. 1 installed on the U.S. Gulf Coast, with
and therefore a predicament arises with multipliers included in the report for
Figure 5 results. A partial resolution of conversion to specific remote areas
this predicament will be illustrated in this throughout the world.
paper by Figure 7. At the same time, if Costs are provided in 1990 dollars.
one considers that only large A single train equals a world class
developments are economically installed in plant size.
environmentally harsher areas of the Sweet, dry gas is the feed to the plant.
North Sea, proper correlations to history 10 days of output = storage capacity.
are depicted by Figure 5. Adding a Costs encompass facilities to storage
variable to the equation should remove only; transportation costs are
some of the bias introduced by excluded.
environmental conditions. Land is bought at $2000/acre.
Engineering starts with conceptual
engineering.
REPORTING AND UPDATING Offsites include only facilities to
RESULTS support the process plant.
An allowance of $5MM is included for
Frequent reporting and updating of results infrastructural costs.
enhances the opportunity for improving Sufficient fresh water is available.
the accuracy of this approach to Appropriate port facilities are nearby
estimating. At the same time it is (within 50 miles)
extremely important to consider the phase Latitude < 60°, Plant site at < 2000
in an exploration/development program. feet elevation.
48
 SPE 2201'

Onshore/non-swampy site. REFERENCES AND INFORMATION

Excludes construction of ports and SOURCES
harbors.
Excludes non-construction housing.
Political interferences and stoppages 1. Albright, S.c., Statistics for
due to strikes are excluded. Business and Economics,
Plant site is already accessible by road. MacMillan Publishing Company,
Imported materials of construction and New York, NY, 1987.
labor, if competitive, can be used to
build the plant.
Severe environmental restrictions are
excluded.
A better method of creating an estimating
technique for developing capital costs in
the screening stages of economic
evaluation is to add data rather than
narrow the scope. In almost every
organization exploring for oil and gas,
numerous estimates of facilities costs have
been made. Merely add results of those
independent estimates to the cost data
presented in Figures 1 through 5, thus
achieving a better correlation. While
uncertainty is introduced by using
estimates containing inaccuracies,
variances due to inaccuracies among
estimates are offsetting to the point of
being insignificant. Using the set of cases
in Figures 3 through 5 consider narrowing
the region to Northwest Europe, adjusting
to permanent facilities only as before,
then adding estimates to the data set.
The resulting curve fit would then give an
estimating approach for a geographical
area larger than the United Kingdom
sector of the North Sea. Figure 7 depicts
the resulting correlation which approaches
providing an "estimate of estimates." In
summary, Table 4 provides the outputs of
all aforementioned comparisons.

CONCLUSIONS
This paper has presented a timely, yet
satisfactory approach to estimating
facilities capital costs for preliminary and
conceptual economic evaluations of oil
and gas developments. Parameters, such
as reserve size, water depth, and
geographical location, were identified as
critical to cost estimates. Finally, one
must be aware of the fact that estimates
are always wrong (it's a matter of degree
or accuracy) but they can be beneficial
tools to successful economic evaluation for
developing oil and gas reserves worldwide.
49
 ·RE 22015
1. _ _
---Prcjoct -~~
..--- -
TABLE 2

._a-- _1_
2._~

TABLE 1
4.
3. CCUrIIy:
External Data Resources e-a. .......
5. _ _ (11)
Libraries Periodicals 6._(11)
..
--------
7. #WIIII 6."_ "".

-
API Library in Washington, D.C. Oil & Gas Journal 1o...... Tana 11. PIe TOIII

- - -
SPE Library in Richardson Ocean Industry 12. DIdc Ten 13.F.-yTCInI
Sea Technology '4ll.CloI(IIl:I) _ _ _ _
Library of Congress, D.C. 14. . . . . . . '''''OI(nwntItII)
Petroleum Economist
'''IlII91~
__ 'Ild._
AAPG Bulletin '!lL CloI_ _ _ '5c. ee.-
Offshore
PoIIc .10.01
'10.01-- _ 'llg. ceraor-. -- ,5h.- --- ___
,........
IlII91 .51. CloI _ _
'Ilb. _ _ _ _ _
Petroleum Review

,. -
Services ,ea. T;;;;;;:- 'ea._ ---
,.....
NorOi!
Wood Mac
James Capel and Company
Petroleum Abstracts
World Oil-International Edition
Petroleum Engineer International ._lIloItOdl
.7.N'I

,..
____ , .. _

I#C
DraIg _(11) _ _ _ _

. ... U!
Arameo World
Geophysics
2O. ~a.
_ _ _ DoIo

....
"'"- ....
.,.ean...-.a.

-.... -
SPE Journal of Petroleum Tech.
........ ........
-
"0. 21e.
World Bank SPE Drilling Engineers
Pipeline Industry
22.~\IIdIX*ll<Y-l
Zl. - . DraIg (-lord)
"".
Notwegian Oil Review
Conference and Symposium Papers
.e-(Il'III
...
25. _ _ ........ I#C

....
U!
....
........
Reference Books 27._
26.1_ .... ....
North Sea Facts ....
27.a

........
Z1Il.
........
27c.

- -
26.T_/T_
North Sea Field Development Guide
International Petroleum Encyclopedia "'lIgI/~
3CL lklb· T....
.... ....
European Continental Shelf Guide
31.~
32.T.... _
........
310. 31b.

"". ........
3'e.

33.
_._a.
Dralge- "".
34. FIIId U. .... - . 37. _ _ _
35b. _

3B.Dodc_·_1OOI)
39. _
3B.CrudO __CO2
TyPo _ ... 3llb._..... --lY/ftl

'"o
TABLE 3
3llO.eomor-
(yarn)
... _ _ Typo
--
--.0;;:-_391._
-- _ _ T_ ---SIllnogo
3llb.~
- 39C.Qwnn

Reduced Set of Project Parameters :=:

I 41. PIpelIne 4'a. Type
142. _ T_I_("/41) -- ---
_OnSltln--...:oi"Sltln
--.'C.L.ength --.,4. ShorI--.,.. OIShcn -
41b. SlzI
-- -
.... 1.onQIlt
On

143.. - LJtlY/n_I___
4S.P1af'II.......... RICOY. _ _ _ _
4B.SIllnogoT_
44&Secanr:tary _
44b. _0_ _ _
TertWy Reccw. _ _ _ _
Country
41. 0peraIing CoIUI (S/bb1) 471. PIanMd ___ "7b. AcIUII

Project Type (Le., Offshore vs Onshore) 48. Data 8cUCIS: c.d:abm

Water Depth

Recoverable Oil Reserves TABLE 4

OurPlIT OF REGRESSION ANALYSIS
Recoverable Gas Reserves
(Y = e"xh)
Total Equivalent Reserves 1.3 x
Coemcient Number Sl.1.ndard
CoerftcienlS of determination of Data Error as:l
Production Capacity
~ :EW!r!t • h _ _r ' _ eilln.!! ~

Worldwide t 0.749 0.954 0.779 9S 210

Implementation Period (Midpoint of Capital Commitment)
Southeast Asia 2 0.409 0.981 0.930 33 100

Reported (or Estimated) Final Capital Cost Northwest Europe 3 0.955 O.99t 0.682 50 ISO

North Sea 4 0.879 1.010 0.696 46 160

Distance to Market
Fixed Platforms
in the UK North Sea 5 0.205 1.195 0.865 '6 tOO
Number of Producing Wells
Northwest Europe
+ Estimates 7 0.759 1.IJ20 0.724 56 t50
Project Adjusted Cost

•Approximates variability consistent with an eighty percent confidence limit.

 seE 2201 5

10,000

i .'
-/
i
~
:Ii
1,000

•....
.'.
/. ~
:Ii
1,000 !:---+-----+--,.'--7'-f-7-----j

~ ~

§ Iii
8
c c 100 !:----h--,"-;f+".:---I------j
100
~:::l ~:::l
~ (3
<
0W 0w
a 10 aa::
If Q.

.
:
~.,

1
1 L..:..L.L.LLUllL-~LLLLWL--L...LLLUllL-L.L.LllllJJ

1 10 100 1,000 10,000 1 10 100 1,000 10,000

TOTAL RESERVES (MMBOE) TOTAL RESERVES (MMBOE)
Figure 1 Figure 2
PROJECT ADJUSTED COST VI PROJECT ADJUSTED COST VI
TOTAL EQUIVALENT RESERVES TOTAL EQUIVALENT RESERVES
WORLDWIDE SOUTHEAST ASIA

10,000 10,000

i i 1,000
~
1,000
~
:Ii :Ii
~ ~
Iii
§ 8C
c 100 100
~
:::l
~
:::l
(3
~ <
0
w
0
w
a 10 a
a::
10
If Q.

10 100 1,000 10,000 10 100 1,000 10,000

TOTAL RESERVES (MMBOE) TOTAL RESERVES (MMBOE)
Figure 3 Figure 4
PROJECT ADJUSTED COST VI PROJECT ADJUSTED COST VI
TOTAL EQUIVALENT RESERVES TOTAL EQUIVALENT RESERVES
NORTHWEST EUROPE NORTH SEA

51
 SPE 2201 5
+100%

i
~
:::lE
~

§
100
m
..,
::l r 1/ / ....
FINAL
5i! SCREENING CONCEPTUAL COST

I.. co:~ .1
ESTIMATE
~ (OME)
a
II:
10
11.
-50% FOR
DEVELOPMENT
ENGINEERING
EVALUATIONS STUDIES
10 100 1,000 10,000
TOTAL RESERVES (MMBOE) AFC FOR DRILLING
Figure 5
PROJECT ADJUSTED COST va 1 DAY 1 WEEK 1 MONTH MONTHS
TOTAL EQUIVALENT RESERVES
FIXED PLATFORM· NORTH SEA (U.K. SECTOR ONLY) Figure 6. Estimate Timing and Accuracy

(II
I\)

10,000

i
~ 1,000
:::lE
~
I-

8
100
m
::l

~
0w
a 10
If

10 100 1,000 10,000

TOTAL RESERVES (MMBOE)
Figure 7
PROJECT ADJUSTED COST va
TOTAL EQUIVALENT RESERVES
NW EUROPE ACTUALS AND ESTIMATES