# Full text of "Completion cost trend analysis."

## See other formats

```COMPLETION COST TREND ANALYSIS

A Special Research Problem
Presented to

The Faculty of the Construction Engineering

and Management Program

Purdue University

by
John T. Baker

In Partial Fulfillment
of the Requirements for the Degree of
Master of Science in Civil Engineering

July 1990

>pxc

Approved:

v

T247832

ABSTRACT

J. Gordon Davis's article "Keeping Project Costs in Line"
( Machine Design , December 1976) stated that managers who
keep track of a project by analyzing cost reports are behind
what is actually occurring on the site. Davis suggests the
use of projected completion costs to analyze the project's
cost control. He states that this method will better serve
the project managers and allow them to respond to problem
areas before they escalate. This paper will analyze Davis's
approach through the use of simulation to determine if this
method is reasonable in the construction industry.

ABSTRACT i

LIST OF FIGURES iii

LIST OF TABLES iv

LIST OF APPENDICES v

CHAPTER I INTRODUCTION 1

1.1 PROBLEM AND PURPOSE 1

1.2 OBJECTIVE AND SCOPE 2

1.3 REPORT FORMAT 2

1.4 STUDY METHODOLOGY 3

CHAPTER II CURRENT FORECASTING SYSTEMS 5

2.1 UNDERLYING THEORIES OF FORECASTING 7

2.2 METHODS OF FORECASTING 9

3.1 SPREADSHEET DESCRIPTION AND COLUMN BREAKDOWN . . 17

3.2 THE FORMULA 25

3.3 SIMULATED RESULTS OF TYPICAL SITUATIONS 26

3.4 SENSITIVITY OF THE WORKSHEET 4 2

3.6 MACRO PROGRAM FOR EXCEL 66

CHAPTER IV RESULTS 67

CHAPTER V RECOMMENDATIONS 69

CHAPTER VI CONCLUSIONS 7

REFERENCES 7 3

APPENDICES 74

11

LIST OF FIGURES

Figure 1.1 - Study Methodology 4

Figure 2.1 - Productivity Profile (Clark & Lorenzo, 1985) . 11

Figure 2.2 - Trend Curve (Riggs, 1987) 16

Figure 3.1 - Cost Trend Chart Worksheet 18

Figure 3.2 - Project Cost Trend Chart 19

Figure 3.3 - Original Spike, Increase in the Estimate ... 30

Figure 3.4 - Original Spike, Decrease in the Estimate ... 32

Figure 3.5 - Steadily Decreasing Estimate 3 4

Figure 3.6 - Steadily Increasing Estimate 3 6

Figure 3.7 - Double Spike 38

Figure 3.8 - Increase Spike with a Following Decrease Spike 40
Figure 3.9 - Sensitivity Analysis, Column 8 Factor Set at

1.0 43

Figure 3.10 - Sensitivity Analysis, Column 8 Factor Set at

3.0 46

Figure 3.11 - Sensitivity Analysis - Cost Slope Factor

Changed to the Power of 4 50

Figure 3.12 - Sensitivity Analysis - Cost Slope Factor

Changed to the Power of 1 5 3

Figure 3.13 - Sensitivity Analysis - Cost Slope Factor

Analysis 56

Figure 3.14 - Improvement Analysis 63

in

LIST OF TABLES

Table 2.1 - Variance Analysis 10

Table 3.1 - Results of the Sensitivity Analysis 59

IV

LIST OF APPENDICES

A. COST TREND CHART WORKSHEET OPERATION MANUAL

B. DAVIS'S ARTICLE

D. EXCEL SPREADSHEET AND CHART MACRO

CHAPTER I INTRODUCTION

1.1 PROBLEM AND PURPOSE

In 1976 Gordon J. Davis wrote an article in Machine Design
which described a method to predict project completion costs.
Davis established a spreadsheet with accompanying chart that
showed the relationship between the estimated cost at completion
(ECAC) with the spreadsheet generated predicted cost at
completion (PCAC) . This method was to be used by project
engineers to identify cost control problems in a project. Davis
stated in his article that by analyzing periodic cost reports,
the manager is at a disadvantage in that the reports are always
lagging behind the actual progress of the project. His entire
premise is that a new method must be instituted so that future
costs can be forecasted and efforts undertaken prior to the
escalation of a negative situation.

The goal of this research project is to determine if Davis's
method is reasonable. A reasonable cost control method would be
evaluated on practicality, usefulness, ease of use, and
manipulation and customization of the method. In addition to the
practicality of this method, the method is also compared to other

current forecasting methods. The method is then reviewed to
determine what changes are necessary to customize the spreadsheet
to a users various requirements.

1.2 OBJECTIVE AND SCOPE

The objective of this research paper is to determine the
usefulness of Davis's worksheet formula in construction projects.
First an analysis of Davis's worksheet and chart is presented
then the basic equations of the spreadsheet are determined. This
report subsequently conducts a sensitivity analysis to observe
the results of the spreadsheet given different conditions.
Finally the spreadsheet and the sensitivity analysis are compared
to determine if the spreadsheet can be improved.

1.3 REPORT FORMAT

This report begins with a detailed introduction to the
problem and states the purpose for this report. Next, a
comprehensive collection of currently used trend and forecasting
methods are presented in Chapter 2. Chapter 3 is the main focus
of the research project and provides an explanation of the
several different ways to improve the spreadsheet. Chapter 4
summarizes the results of the research and Chapter 5 provides

recommendations for the use of the spreadsheet and also suggests
follow on research. Finally the report is concluded in Chapter
6.

1.4 STUDY METHODOLOGY

As can be seen in Figure 1.1, the study of Davis's
predicated cost completion method begins with an analysis of the
individually and then the formulas that are used within the
spreadsheet are determined. A sensitivity analysis of the
spreadsheet is performed to show the reaction of the predicted
cost at completion (PCAC) when given the estimated cost at
completion (ECAC) . When reviewing these charts if must be
understood that the ECAC is the best possible estimated cost at
completion that the project manager can estimated after reviewing
all available information. The PCAC is the predicted completion
cost at completion provided with the assistance of Davis's

Finally the sensitivity analysis and the formulas are
compared to develop an improved spreadsheet.

THE PROBLEM:

I

REVIEW AND DESCRIBE

I

BREAKDOWN AND
ANALYZE EACH COLUMN

i

ESTABLISH THE EQUATION

I

PERFORM A SENSITIVITY
ANALYSIS

Figure 1.1 - Study Methodology

CHAPTER II CURRENT FORECASTING SYSTEMS

In a typical construction project, the company cost
estimator develops an estimate which reflects all costs that are
required to complete a specific project. These costs are usually
broken down into specific cost categories which ease in the
development and periodic monitoring of the project. The
development of the estimate or budget is produced by a
knowledgeable and experienced estimator. The budget is
continuously reviewed to ensure accuracy and to minimize
omissions. This estimate is extremely important, the company
profit margin and ultimately it's life expectancy rests in the
hands of the estimating staff. If the estimate is too low, large
cost overruns might occur, if the estimate is too high, the
company will lose business in the competitiveness of the
construction industry.

Throughout the construction of the project, the project
manager periodically reviews costs incurred and compares this
data to the budget. Frequent monitoring of these costs are
essential to control expenditures and insure that costs do not
overrun the budget. This process of continuously monitoring,
managing, reacting to variances and forecasting is termed cost
control .

Every project manager has a strong incentive to monitor

costs on the project. The company profit is directly related to
how well the estimated budget was developed and the manner in
which the expenditures were managed.

In this chapter, several different forecasting methods are
presented. These methods are presented so that a comparison can
be made between what is currently being used in the industry and
what Davis's formula represents.

2.1 UNDERLYING THEORIES OF FORECASTING

In the past, cost reporting and control was performed by the
on site field superintendent who also performed the forecasting
of future costs. This function is now in the hands of the
project manager due to the complexity and importance of cost
control.

Forecasting is the process of predicting the future by using
the analysis of all available data trends to ultimately calculate
the final results reguired. In the construction industry this
ultimately results in determining the final construction cost of
the project through the analysis of past trends, experience, and
any and all available data.

Forecasting in the construction industry can be broken into
three separate theories, they are:

1. The performance of the construction project over the
balance of the project will follow the estimate/budget.
Therefore the projects overruns or underruns will remain the same
for the rest of the project as stated on the report date. For
example, if the project is currently at 35% with a cost overrun
of \$3,000, this forecasting theory states that the remaining 65%
of the project will not have any variances, that is no additional
cost overruns or underruns and therefore the completion cost will
have an overrun of \$3,000.

2. The performance of the construction project over the
balance of the project will follow the overall past performance
prior to the report date. For example, consider construction of
a highway with a budget unit cost of \$25,000 per mile, 12 mile
project, for a total budget of \$300,000. The current report
shows 5 miles complete at a cost of \$100,000, which is an average
unit cost of \$100,000/5 = \$20,000 per mile. Therefore the
forecasted completion cost of this project would by (\$2 0,000 * 7
miles) + \$100,000 = \$240,000, a total cost underrun of \$60,000.

3. The performance of the balance of the project is
determined by the analysis of the productivity levels at the time
of the report. Using the example in item 2 above, say the last
mile completed, mile 5, cost \$19,000 to construct. The
forecasted completion cost of the project would then be (\$19,000
* 7) + \$100,000 = \$233,000, an underrun of \$67,000.

8

2.2 METHODS OF FORECASTING

Forecasting methods require a detailed cost summary produced
periodically. These cost reports are usually produced on a
monthly basis with each work item is broken down into a separate
cost category for the purpose of recording and analyzing. Cost
codes are frequently used and some companies have customized
these codes to their specific requirements.

The monthly updates are a detailed accumulation of actual
costs incurred as of the report date. These costs are then
cateqorized into the specific cost codes and reported through a
computer system.

The methods described in this section are a representation
of currently used methods throughout the construction industry.
Some of these methods have limited uses, some of the methods can
be used in combination with other methods and others are
customarily used by themselves. The optimum combination of the
methods must be customized to the individual company
requirements. But even with the best combination of the methods,
the ultimate producer of the forecasted figure must apply
experience and common sense to the final results. It should also
be noted that many forecasting methods are not used until 2 to
25% of the project is completed (Bessa, 1983) .

A. VARIANCE ANALYSIS

Variance analysis is the simple process of comparing the

current performance with the expected performance. Typical items
compared are:

1. Budget Cost vs Actual Cost

2. Budget Unit Cost vs Actual Unit Cost

3 . Budget Manhours vs Actual Manhours

4. Budget % Complete vs Actual % Complete

This method is not meant to be used by itself but other
contributing factors must be analyzed as well to obtain the whole
picture. This method highlights cost categories that are not
performing or are performing better than anticipated. To use
this method, contributing factors such as the following must be
scrutinized to affirm that there is trouble:

1. Poor initial estimate

2. Technical difficulties

3. Unexpected labor or material costs

4. Differing labor efficiencies.

Forecasting is easily computed using any of the three theories
above with the available cost data. Examples are provided in the
previous subchapter and in Table 2.1 - Variance Analysis.

Cost
Code

Expenditures

Costs to Date

Estimate to ComDl»t»

To Date

Quant.

Unit

Amount

Quant .

Unit

IO

CtW jcT fcXc Av'

657'

6 c/ -

loe>0

o.kS

4

4

ii

l ~vo<>> B<C'-tf""-<-

Colo "

?o~h>-

500

4.14-

i

—

>1

A<£« BA«fc -o^i

14 2- SO

az.n>

°\°

1 So

t

—

Forecast of Final Cost

Original Estimate

Amount

% '

Amount

Quant.

Unit

Amount

Quant.

Unit

Over

Under

6ST

|ooo

o-W

bOO '

\fSBt>

o fc

5"!"

fo&To

7t>*>-

SbO

4.i+

7Dtt>~

9*>

4-OD

Ml"f«

\4i-a>

to

ifco

-LkO-

°I0

^o\>

!?•*>

Htfl.

Table 2.1 - Variance Analysis

10

B. PRODUCTIVITY PROFILES

Experience shows that productivity does not remain constant
throughout the term of a project but follows well established
curves and patterns called productivity profiles. The typical
curve begins with poor productivity rates and then increases to
reach a maximum at the 30% to 80% range before tapering off to
final completion. This slow start is sometimes called the
learning curve of the process in that workers take time to
organize and orientate themselves prior to becoming effective.
The final 20 - 30% of the project is also a time of weak
productivity as crew composition changes, rework and cleanup take
an increased priority.

In the analysis of direct labor, the productivity profile is
one of the better methods of trend analysis and forecasting
available. The profile is a graphical representation of the
productivity rate vs the % physical completion, Figure 2.1 -
Productivity Profile. Productivity is defined as the Budget Unit
Rate divided by the Actual Unit Rate.

% Budgeted Manhours Used

Theoretical 9.7 18.2 26.3 35.4 44.6 54.5 64.2 74.8 86.5 100
Actual 10.4 18.7 28.3

1.2 -i

1.1 -

- 1.0 ,S S

tj 0.9 -/ N

3 ' ACTIIJ

2 0.8 -

B.

0.7 -
0.6 -

CALIBRATION CURVE

\

ACTUAL

-I 1 1 1 1 1 —I 1

SO 40 50 60 70 80 90 100

Physical Completion (X)

Figure 2.1 - Productivity Profile (Clark & Lorenzo, 1985)

11

Forecasting in this method involves the interpolation of the
productivity profile with the latest cost report. The method
begins with the comparison of the actual productivity rate and
the estimated productivity rate. The difference is used to
calculate the forecasted final productivity rate. For example,
suppose the estimated productivity rate is 1.12 and the actual is
1.06 the difference being 0.06. The % difference of the rates is
0.06/1.12 = 5.4%. The final productivity rate would be 1.0-0.054
= 0.94 6. Predicted manhours for the entire cost code would be
50,000/0.946 = 52,854 manhours.

C. MANPOWER PERFORMANCE FACTOR

The manpower performance factor is a method of forecasting
which relies on the theory that the future performance of a
project can be measured by the average past performance of the
project. The manpower performance factor is defined as (Riggs,
1987) :

ESTIMATED MANHOURS TOTAL-TO-DATE QUANTITY
MPF = *

ESTIMATED QUANTITY TOTAL-TO-DATE MANHOURS

The manpower performance factor is the same as the Cost

Performance Index which is defined as:

EARNED MANHOURS OR DOLLARS
CPI =

ACTUAL MANHOURS OR DOLLARS

12

where ,

ESTIMATED MANHOURS

EARNED VALUE = * ACTUAL QUANTITY

ESTIMATED QUANTITY

The MPF factor is simply a ratio between the estimated rate of
production and the actual rate of production. The estimated
manhours to complete the project is therefore the MANPOWER
PERFORMANCE FACTOR multiplied by the remaining quantity. The
total projected manhours to complete the project would therefore
be the total of the estimated manhours to complete plus the
manhours to date. Projected cost of the activity would then be
the estimated manhours to complete the task multiplied by the
actual cost per manhour added to the cost incurred to date.

An example of this technique is described below (Riggs,
1987) :

GIVEN: 1. ESTIMATE:

Quantity: 4,000
Manhours: 3,080
Cost: \$29,280

2. TOTAL TO DATE:

Quantity: 3,990
Manhours: 3,112
Cost: \$29,875

MPF= 3080/4000 * (3990/3112) = 0.99

ESTIMATED MANHOURS TO COMPLETE = MPF * REMAINING QUANTITY

= 0.99(4200-3990)
= 2 08 MANHOURS

PROJECTED MANHOURS = ESTIMATED MANHOURS TO COMPLETE * MANHOURS TO

DATE
= 208 + 3112
= 3 32 MANHOURS

13

ESTIMATED COST TO COMPLETE = ESTIMATED MANHOURS TO COMPLETE *

ACTUAL \$/MANHOURS

= 208 MANHOURS (9.60)
= \$1,997

PROJECTED COST = ESTIMATED COST TO COMPLETE + COST TO DATE

= \$1,997 + 29,875
= \$31,872

D. COST PERFORMANCE INDEX

This method is similar to the manpower performance index but
instead of using manpower productivity ratios this method
utilizes the cost index. This method compares the differences
between the estimated cost of the activity and the actual cost of
the activity, this ratio is then applied to the remaining portion
of the contract to determine the forecasted completion cost. For
example (Riggs, 1987) :

BUDGET COST TO DATE: 7650
ACTUAL COST TO DATE: 8 37
TOTAL BUDGET COST: 21,000

CPI = BUDGET COST / ACTUAL COST = 7650 / 8370
= 0.914 < 1.0 therefore unfavorable

BUDGET COST OF REMAINING WORK = 21,000 - 7,650 = 13,350
ESTIMATE TO COMPLETE = 13,350 / 0.914 = \$14,606
FORECAST AMOUNT = 8370 + 14606 = \$22,976

14

E. TREND CURVES

Trend curves are curves plotted with the horizontal axis as
percent complete and the vertical axis of manhours, cost or other
productivity factors, Figure 2.2 - Trend Curve. It should be
noted that this graph can be plotted directly from the
productivity profile. By plotting dates with the percent
complete the schedule can be integrated into the cost reporting
aspect and a comparison can easily be made.

Forecasting with the trend curve is best described by the
use of an example (Riggs, 1987) :

1. ACTUAL MANHOURS COMPLETED: 43,000

2. LESS THEORETICAL MANHOURS

AT 30% COMPLETE, (from the calibration curve) 40,700

3. OVERRUN MANHOURS 2,3 00

4. % OVERRUN = (2, 300/40, 700)*100 5.7%

5. TOTAL ADJUSTED MANHOURS 12 0,000

6. OVERRUN MANHOURS = 0.057(120,000) 6,800

This method assumes that there is no action taken to correct the
problem and that the problem continues to compound at the rate
calculated. A second assumption is that the manhours will remain
a constant percentage relationship with the calibration curve.

15

Figure 2.2 - Trend Curve (Riggs, 1987)

16

Davis's article as shown in Appendix A describes a
completion cost worksheet. This worksheet requires several
inputs including the estimated cost of completion (ECAC) , and
determines the predicted cost of completion (PCAC) . This chapter
reviews the formation and column breakdown of the worksheet, the
inputs of the spreadsheet, the formula used to calculate the
understanding of the spreadsheet different scenarios were
developed to analyze the cause and effects of several situations.
From these scenarios the sensitivity of the worksheet is analyzed
and finally several improvements are recommended.

3.1 SPREADSHEET DESCRIPTION AND COLUMN BREAKDOWN

Figure 3.1 and 3.2 show the worksheet and the graphical
representation of the results, called the Worksheet Chart. The
top portion of the worksheet is set aside as the title. The
worksheet consists of 12 numbered columns that are labeled on the
top of each column. A title of each column and the formula for
the column is also stated above the numerical data. Reference
columns at the beginning and end of the worksheet are used for

17

.rci.u •«. .jriKj-ji-.cr ;

.... '.'.■.'.■.'.'.■.'.'.'.'.'.■..■. .'-..■ V.Y. '.".". - .'. . .*.' .■...'■ V. .".".'- . .V.'.Y. .

DA VIS ORIGINAL ARTICLE

ORIGINAL COST= : 5

4 ■". - i C

1 i. j *t -J j

PREVIOUS i PROJECTED REMAINING ESTIMATED

r.r-r-..— rf-.-r n r- r-. .— . rvr r*. r- r, .— . r.-r » ir.r. * tt" nr.kjni n"xii-".ki Tikjr r-nr-T * *r

ncruni niruni rxcrum : jru.-.ic uunruLiiun uric luoi .'-. . ;

NUMBER DATE DATE : INTERVAL ; DATE i TO COMPLETE ; COMPLETION

ikii-ii rr .— —.i t .—-<—.» •» ikini t-r

ECAC

|NAL 8 5.00

0RT1 10 16 5 5.00

phi - i i oo y.ju

ORT 3 3 2 1-3 5 5.50

UPi 4 4 j 1 loud J.UJ d.Ud

C iC T" 6 4 2 3.77 3 77 8 50

.—i rvr .- -j r * r. r -? n ,- -j ,-• r. r.
■_;.-<. I i l J.Oi ^.01 O.iJV

OPT? 8 7 1 9.75 1.75 9.00

.—. r-.-r .-. .-. a r.

'_.- i ; o o U "0

ORT 3 9 3 -9 !

.— ,r.T * .-. < .-. ,-. < < .-.

Ljn. i iu :u o i -iu

i

: : i
: :

; ; ; i

. i

! i

8 3 JO 11 12

COST SLOPE ADDiTlONAL COST aT REPORT

nycpoMW OVERRUN ; COMPLETION NUMBER

S-ORIGCOST 2(3x7] SUMS 9/1 "2 5x1 2 6+11

j : __j _ ;

;j ijij f; Qf) ij Qlj i'i mm j mm c , "if: REF'jRT 1

50 i 00 1.00 0.25 1.50 7 00 REPORT 2
50 1.00 2 00 22 111 6 61 REPORT 3

1.00 2.00 4 00 0.25 2.27 8 27 REPORT 4
■ 50 MOO 18.00 0.50 1 ''"' J 10 " J 9 REPORT 5

•*00 8.00 26 00 58 1.42 10.42 REPORT 6
4.00 8 00 3 J 00 0.53 0.33 8.33 REPORT 7

-5.00 00 34.00 0.53 -4.25 REPORTS

-5 00 00 34 00 42 -'■ 73 REPORT i

-5 00 -10.00 24 00 0.24 -2.40 REPORT 10

I

12.00

INDEPENDENT RESEARCH PROJECT

DAVIS ORIGINAL ARTICLE

10.00 +

_n

"fl

r> on

o.uu

n

.4 nn _L

? nn -L

rL

-ir"

'tJ

! -I- ECAC !

! I

r __ . _ i
I -U- PUAU i

n on

3 4 6 7

MONTH

8 8

10

.

.

reference of the specific rows.

The chart shown in Figure 3.2 is a graphical representation
of the two curves, the ECAC and the PCAC. The horizontal axis
represents the update interval, usually in months. The vertical
axis represents the completion costs. With every chart provided
in the report, both a title and chart key are added.

The most important columns of this worksheet are the input
column number 6, the ESTIMATE COST AT COMPLETION (ECAC) and the
resulting PREDICTED COST AT COMPLETION (PCAC) column 12.

The following is a column by column breakdown and

INITIATION OF THE WORKSHEET

TITLE: this space is provided to input the title of the
construction project and any pertinent comments.
ORIGINAL COST: the original estimated cost at completion is

inputed into this worksheet block.

REPORT NUMBER

Prior to column one, the worksheet presents a column showing
the report number. In the analysis of this worksheet these
updates are stated in monthly intervals. However the
updates may be in any interval as long as the unit of the
interval is consistent.

20

COLUMN 1: REPORT DATE

This column is a restatement of the previous column. The
value of this column is used later in the spreadsheet, the
word "REPORT" is therefore eliminated in order for the value
to be used in the spreadsheet program. Similar to the first
column this column shows the report date as a monthly
update.

COLUMN 2: PREVIOUS REPORT DATE (INPUT)

This column simply states the last report date and is
inputed as the number of the last report date.

COLUMN 3: UPDATE INTERVAL

The update interval is the difference between the first
column and the second column. This shows the amount of time
that the updating of this spreadsheet has lapsed.

COLUMN 4: PROJECTED COMPLETION DATE (INPUT)

This value is inputed into the spreadsheet as the estimated
time to compete the project. Note that this value is part
of the periodic updating of the spreadsheet. This value
represents the total duration of the project, not the
remaining completion time.

21

COLUMN 5: REMAINING TIME TO COMPLETE

This column is automatically calculated by subtracting
COLUMN 1: REPORT DATE by COLUMN 5: PROJECTED COMPLETION
DATE.

COLUMN 6: ESTIMATED COST AT COMPLETION (INPUT) (ECAC)

This input value is the most up-to-date completion cost
estimate of the project. Usually a project manager uses a
combination of the methods described in the previous chapter
to determine this value.

COLUMN 7: ESTIMATED COST OVERRUN

Column 7 represents the estimated cost overrun for the given
interval. This calculation is simply the difference between
COLUMN 6: ESTIMATED COST AT COMPLETION and the ORIGINAL COST
as imputed during the initiation phase of the setup.

COLUMN 8: CALCULATIONS

This calculation consists of multiplying the product of
COLUMN 3: UPDATE INTERVAL and COLUMN 7: ESTIMATED COST
OVERRUN by a factor of 2.

22

The reasons why the factor of 2 is applied is not apparent,
in the sensitivity analysis section of this chapter this
factor is discussed.

COLUMN 9: CALCULATIONS

This column is the summation of the previous column to the
point of the update. For example, if the report date is
number 5, then the value of this column would be the
summation of the first five entries of COLUMN 8 .

COLUMN 10: COST SLOPE

The cost slope is derived by dividing COLUMN 9 by the square
of COLUMN 1: REPORT DATE. As can be seen from the graphical
representation of this spreadsheet, the cost slope is the
slope of this line between the two update intervals.
Without the square root of the denominator this equation is
basically the average of the interval cost overruns.

In Chapter 3.4 Sensitivity of the Worksheet, a discussion of
why the denominator is squared is provided.

This value is calculated by multiplying COLUMN 5: REMAINING

23

TIME TO COMPLETE by COLUMN 10: COST SLOPE.

COLUMN 12: PREDICTED COST AT COMPLETION (PCAC)

The final column represents the results of this spreadsheet
and is calculated by the addition of COLUMN 6: ESTIMATED
COST AT COMPLETION and COLUMN 11: PREDICTED ADDITIONAL
OVERRUN .

24

3.2 THE FORMULA

The formula used in the spreadsheet can easily be broken
down into two equations. These equations are:

PREDICTED COST = ECAC + REMAINING TIME * COST
AT COMPLETION TO COMPLETE SLOPE

COST 2 (UPDATE INTERVAL) (INTERVAL ESTIMATE COST OVERRUN)
SLOPE = (REPORT DATE) A 2

Two most important factors in these formulas are:

1. the square of the report date in the denominator of the
cost slope, and

2. the factor of 2 in the numerator of the cost slope.
Both of these factors will be discussed in detail during the
worksheet sensitivity and worksheet improvement sections of this
chapter.

25

3.3 SIMULATED RESULTS OF TYPICAL SITUATIONS

To gain a better understanding of the spreadsheet and
equations, several different scenarios were developed to show the
cause and effect of the way in which the spreadsheet functions.
The parameters of these simulations were held constant to enable
comparisons of the different situations. The parameters for this
simulation consisted of the following:

1. The time frame is held constant at 10 months.

2. The original cost is 20.

3. All changes in the ECAC deviated from the original 20.

4 . Every worksheet was updated constantly and no reports
were omitted.

All of the spreadsheet results are shown graphically on
the accompanying chart. The graphing of the results allows the
analysis of the different trends presented between the ECAC and
the PCAC. In the graphs the horizontal scale represents the
periodic intervals, in most cases this is on a monthly basis.
The vertical scale represents the costs of the completed project.
The two lines within the graph represent COLUMN 6: ESTIMATED COST
AT COMPLETION referred to as the ECAC and COLUMN 12: PREDICTED
COST AT COMPLETION referred to as the PCAC.

A. ORIGINAL SPIKE WITH FOLLOWING CONSISTENT ESTIMATES

1. INCREASE IN ESTIMATE: Figure 3.3 - Original Spike with

26

Consistent Following Estimates. This scenario was developed
by changing an early ECAC at 20% and continuing this
estimate for the remainder of the project. This would be
similar to a project manager finding a problem in his
estimate at an early date and then continuing with the
revised estimate for the remainder of the project. In this
simulation, the ECAC for the second interval increased 2 5%
but the PCAC increased 125% over the initial estimated cost,
from 20 to 45.

2. DECREASE IN ESTIMATE: Figure 3.4 - Initial Spike with
Consistent following Estimates. This plot was developed the
same way the above "spike plot" was simulated. This plot is
completely unreasonable in that the PCAC drops below zero,
in other words, the predicted cost at completion of this
project is zero. This plot shows one of the major flaws in
Davis's premise, a PCAC less than zero.

1. DECREASING: Figure 3.6 - Steadily Increasing Estimate
The ECAC was steadily decreased at a straight line rate of
5% of the original cost estimate. The PCAC as plotted shows
a drastic drop in the first interval and then differences
between the ECAC and the PCAC in subsequent intervals are

27

not as drastic. This affect is also referred to the
convergence to the PCAC curve to the ECAC as the update
intervals approach the completion of the project. This
would indicate that changes in the ECAC at an early date in
the project are more critical then in the late stages of the
pro j ect .

2. INCREASING - This plot is similar to the preceding plot
except the steady changes in the ECAC are increasing. Again
the profile of a large deviation at the beginning is slowly
minimized by the typical convergence of the PCAC to the ECAC
towards the end of the project.

C. THE DOUBLE SPIKE

Figure 3.7 - Double Spike in the ECAC of Equal Magnitude
details two jumps in the ECAC of equal magnitude at different
times in the progress of the project. These jumps produce
different results. The first jump in the ECAC of 25% (25/20)
resulted in an increase of the PCAC by 64% (32.78/20) whereas the
second jump increased the PCAC by only 24% (26.22/21.11). The
differences between the ECAC in the first jump was 32.78 - 25 =
7.78 or 31% of the ECAC in the second jump the difference was
26.22 - 25 = 1.22 or 5% of the ECAC. This difference of 31%
towards the beginning of the project and the 5% difference at the
end of the project is also a factor of the convergence trait of

28

D. INCREASE SPIKE AND THEN A DECREASE SPIKE

Figure 3.8 - Double Spoke int he ECAC, One Increase and One
Decrease shows the behavior of the PCAC curve due to two jumps in
the ECAC. The first jump is similar to the first jump in the
preceding simulation but the second jump is the same magnitude
however opposite directions. Note that the two curves (ECAC and
the PCAC) converge at the second jump and continue with identical
values throughout the remainder of the project.

29

INDEPENDENT RESEARCH PROJECT

ORiGJNAL SPiKE WITH CONSiS I LN I hOLLGWiNG t3 S iMA I fcS

4b.00

U

40.00 4-

n

25.00

-n
-3

U

-1- ECAC

20.00

i
i

1j

! u- PC AC

i

m nn

5.00 J-

n nn

MONTH

ure 3.T - Original 3|. ike,
; ncrea .e in * i" a Est unai e

lU

Pntje "0

COST TREND WORKSHEET

COST TREND WORKSHEET
ORIGINAL SPIKE WITH CONSISTENT FOLLOWING 'ESTIMATES

■

:•:•:•::•>:■:■:.. |

:
1

.ms i

!
•XvX-X-X-X

j"jUrrb' : j

i

ORIGINAL COST=

20

• i

"k

n

c
y

& i

: PREVIOUS

PROJECTED

REMAINING

ESTIMATED

r.r-r-.rAr-.-r

RcrUKi

htruhi

r.t-r; .-> r.*r

ncruni

i t r> r> .-. tt-

Lji'irLc :iui l i

t inc

r-r-.r-T .. t

juii ,-. i

NUMBER

DATE

DATE

INTERVAL :

DATE

TO COMPLETE j

COMPLETION

INPUT

COL4 -COL1

INPUT

irui * i
Ulri,-,i_

a

•* r.

jj.uu i

|)RT 1

1

i

10

M

20 00

r.,— .r.T --,
"UP.! i.

it

i

1

ID

J

-,c r.r.

iy.yy

5 ORT 3

s

£

1

10

7

25.00

Sherr *

| L h i t

"f

3

i

. u

r-

iy.yu j

ERT 5

c

4

1

10

5

25.00

-,.— .r.-r .-

£

c

J :

i

■i r,
1U

-t

AC AA

-j.ju ;

pRT 7

1

b

1

10

:

25.00

"*.— i^.T A

■uni -j

d

"9

1

: I j

i

AC A A

ij.yy j

pRT 3

3

: J :

i

10

1

25.00

-..—.r.-r -i .-,
" UP. 1 1 U

i A

! U

'j

■i

■* r,

'J

Jj.U'j

">

3

10

■* 4

: 2

ESTIMATED

COST

PREDICTED

PREDICTED

OUO 1

n r.T~.r-

« r-.r-, i-nr~,ki .. i
. W .UU!I 1UI1.-.U

— — . --"P -i T

ntruni

OVERRUN

OVERRUN

COMPLETION

NUMBER

a .— .r»ir* r*rn--r

0" urj'j Luoi

r*l Ik J A

3(1 i

c v ■* r.

O-r ! !

1
i

PC AC |

0.00

0.00

0.00

0.00

0.00

20.00

REPORT 1 !

J.UU

I 'JUL'

10 00

a cr.

i.yy

"'"i r, f:

Vj.yy

nrnnnT -■< I

Rcruhi ^

5.00

10.00

20.00

15.56

40.56

REPORT 3

c .-..-.
y.yy

a n r.r,

ili.uu

a a r.r.
OU.U'J

1 .00

4 4 AC
1 1 .J

i\F- -,C

nmnnT j
rxcrun i -t j

5 00

10.00

40.00

1 30

3.00

33.00

REPORT 5

C .'.A

U.L'U

* r, no
i y.yy

Cfl a a

J'J.UU

< A A

I .JO

c cr-
J.JO

-A cr-

j'J.JO

nrr.rir.T ,-■

ncruni c-

D.Li J

10.00

60.00

1.22

J.O(

23.67

REPORT 7

c .-..1

J yy

< fl A A

1 O.'JU

-ja r.r.

i y.yy

i Li J

2.13

27.13

r.r-r.r-.rvr a

p.cr uni o

5.00

10.00

SO. 00

0.33

0.33

25.33

REPORT 3

C |"|fl

10.00

30 00

30

y.'uO

-y.yU

r. r- n — . r.-r •« .-.

r\trun i . v

Figur::- 3.3 - Original Spike
. .■ Ln t iie 'C ■ ' Lma > e

i 1 ..^^

INDEPENDENT RESEARCH PROJECT

INITIAL SPIKE WITH CONSISTENT FOLLOWING ESTIMATES

20.00

■n

ij.uu "f-

a s

j r

p

10.00

5.00

rr

n--

H

i F FOAC

■U- PCAC

n nn

u.uu

— pr

U

/3

10

-5.00

i, /

fr

MONTH

Figure 3. 1 - ": . Lniil Spike,
reasi Ln tli« Cst im.it e

COST TREND WORKSHEET

COST TREND WORKSHEET
MTfAL SPIKE WITH COHSISTEHJ FOLLOWING ESTIMATES

^v;';';';-;;-;-;^;';;-:-:-:-;-:-:-;-;;';-;';-

LJ.-. IC

-.n_H. -.ft

ORIGINAL COST=

20

i

L

3

4

cr

b I

i PREVIOUS

PROJECTED

REMAINING

ESTIMATED

r,rr..-,nT

rerun

nrnriP-T

ncrUru

r.r-r*./-.r-.-p

ncruni

t inn «-rr-

rr.kifti i — n,-,ki

uui'irLhiiuiN

iir:C

.— ■ /—. r t a -r
UUj 1 ,", :

NUMBER

DATE

DATE

INTERVAL j

date

i TO COMPLETE j

COMPLETION

• k i r*. i rr

,-(-.! 1 .--.-.1 4

LuL-t h^uli

ikir.i it

iiiru i

ECAC

3INAL

10

20.00

0RT1

1

1

10

9

20.00

ur.: 1

1

1

1

10

4 C ft ft

I3.UU

OPT 3

ft

I

1

10

7

15.00

.1— .r.T i

uni n

i

i

1 u

b

i y.oU

5

4

1

10

5

15.00

'OFT S

r-

o

i

1 u

A

1

4 C ftft

1 -.-.UiJ

>0RT 7

7

s

■t

10

3

15.00

•.rtrvr r,

■uni o

-

i

! 'J

J

«'«-'";-..-. !

ly.uu ;

'ORT 3

9

1

10

-

15.00

■ir-.r.-r ^ .-.

■uni iu

4 ■"■

1 U

o

4
1

lli

i

4 C ftft

:0.uo i

i

: : : 1

-j

1 li

1 1

1 £

!
j

ESTIMATED

COST

PREDICTED

PREDICTED

.— - .— . r t

.-1 r~.r-.r-

JLurc

A r-.r-,mr- .k t .. i

■-■.— ..-T /, T

n r- n i — i r-iT

rcruai

OVERRUN

OVERRUN

COMPLETION

NUMBER

ft .— .i-.i.— r--r-..-T
0" UPI'J LJJ 1

i\ Jai i

-i i. a r t

ft U 'ft
O,' 1 i

c. . ■* r.

O+l i

i

PCaC

: ; i i
ill ; I

0.00

0.00

0.00

0.00

0.00

20.00

REPORT 1 i

C ft ft

~0.\JU

■* r, ft ft

"I U.l/U

"1 O.uu

ft Cft

-iU.UO

C ft ft

-j.iju

n r- !-,.-. r-i-r ft 1

rcrun.i ..

-5.00

-10.00

-20.00

_ft 70

-15.5b

-0.58

REPORT 3 i

"5.00

-10.00

-30.00

"1 .00

-11.25

3.75

r*. t- n rt r-iT ■

ncruni *t

-5.00

-10.00

-40.00

-1 .80

-3.00

7.00

REPORT 5

-5.00

4 ft ft ft

: -iu.uu

Cft ft ft

-yu.ull

4 -".ft

"I .00

-5.56

3.-H

r.rrnnT r-

ncruni o

-5.00

-1 00

-bO.OO

-1.22

-o.b?

11.33

REPORT 7

c rift
"V.OU

4 ft ft ft
"lU.UU

-"ft ft ft
"fU.UU

4 ft ft

"1 .'JO

ft 4 ft

~±.l

4 ■"■ !%M

1 i. . 3 i

r-, r- r-. r*. r-.T- .-.

ncruni o

-5.00

-1 00

-30.00

-0.39

-0.93

14.01

REPORT 3

-^ ~*~!

-1 0.00

-30.0U

-0.30

0.0'u

4 C ft ft

! J.UU

nrr.np.T -i r. 1

ricr-jr'.i i 'j

f Lgure 3.4 - Oi . -;. i n. .

Dei rea >e in *iit? lj( Lsuitr-

Page

1

1

O

<

< !

1 o

Q |

1 w

Q- 1

! ■

n i

u

5

f\

i U

i _
T u -'

/ !

4- -x

LU

L_
■cT*

V-
M

LU

I-

1- <°

I

U

t-J

2

s~

LU

_j

,T

Q

a--

<

■a"

LU

I

I—

G

d

4-

4-

/

•-U

LF

a

ca

CO

r,
—

—

l Lgure

j

■ l ■ !. ...

P I ■ I

4

mmm

YyXyXy.Y.Y.Y.Y.

■xxxxyXyXXX

yXyXyX

1

:■»■•:■:■:•■;:::•:■:;■:::

•X;:;X;:;XvX;XvX;X

Ly.y.y.y.y.y. . .•■••
.V.V.Y.Y.Y.Y

.■.•.•.•.•.■.V.V.V.V.V.V

■ • ,v ■ • ■ V

x : x'x-:->x :

i^ilicoia.;

: i= c >
a o

:uj! ;n-

1-
~l

CL

z

n

■5

LU

D

O

d
cm

CO • CD

O

Q

r-

oio

(S3; 10

o

I—

O

>ro

o

c

w

1—

o

o

T—

P"

LL
LLI
CC

LX:
LU:
CO;

_];

z

r-

H

X

o

LU

X

CM

r—
X
i i
X
LU
X

ra

X
Q

UL
LU

X

r-

X

u-

LU

X

UO

X

■X
LU

X

3

1-

X

LU
X

X

LU

X

m

r—

X

LU

X

C5!

H

X

u_

X
X

1

r"
1

cl
01

Ssi

i

j

XXXvXvXX-X-Xv

yXyXyXyXyXx:

: XyX : x : x : x : x : x : x : :

X-X-XvX

:

,,.,, v v
Iy'^Sy'IyXX^

I

•XXXvXXXvXX

• ;

: .-- : : 1—

H. LU
. '^ : : _' .

: Z : ^ : i '

• iiO •' rf ■ & • *r
_i — r i

: 5-: r— : T'.
: Ui : : "

Q

rj

I

T

CO

00 N

■i'

UT tj

'70

■? i

T—

■10

T—

LJ

LJ

o

G

LU
LL
LL

<

1—
CO

r-!r-

LU r-

£ +

Q

fj
<f

n

Cl

G

C-)

a

a
in

r-

P3

o

111

o
a

CM

r-

!--
X

T—
—

00

—

P"

uo

T-

I"

p-

T—

P"

!

i
i

ol

■
r-|
p-j

i

|
_i

:— I

"I

i

i

i
i

i

■XyXXXXXyXX

Xyx;: x :y:

.Y.Y.Y.Y.Y.Y.Y.Y.Y
Y.Y.Y.Y.Y.Y.Y.Y.Y

XyXyXyXyXyXy

mm w

.W.Y.V.Y.Y.Y.Y.Y. -

■• : : :: : : - : : : x ; : : x : : : x : x :
•X- xx-x-x-x

XyXyXv/XvXvX;
[•X-X-XXyX-XvXy

-x : x if

& I

:«Xy:y>»

■ » %
t, * •«

::« |

i , v

t § ft

.■::■:+£
|v»:-x :•>: <

lXyyXyXyXyXyX

i ■■■■■• -x ■■•■
k-x-xx:

^•XyXyX-X-XyXy

L .

r ■

! y>>>>>>>.

! "

o! •

C5:

1 :

g;

ji

mi

f";

:

ui

U H

?0

1-
I

LL

z

O

oio

o

o o

r-:T—

o

T—

o

o

o

r"
r-

1 . 1

H

o

Q

LL 1
LL
n

<

u
Q
<

5!w

fcir-

&:x

o

Q

o

1

!■•-
i

o

iij
i

C
!

m

CO
00

i

N

u"

CM

1

uo

N

i—
i

CD

X

i

'Ji-
h-:

<

,; ;;>

.'.'J: :*:LL
: :Q:i.i

! JCLjP

3Z

T—

r-:? -

T™

l-:l—

I—

-

T—

r-

Q

T—

I—
CO

LU

Cl

CO

:cj

;<

c

Q

in

Q

l
t

r-

X

i
i

uo

d

Q

CD

a
i

CD

d
1

ffl

—
1

CO

X

d
!

00

X
1

i

i

X:

.1

o|

lj
i

i

ico;

Qip-i

n¥r;

X

r- ri

m

■*:in

(0

f--

CO

0)

CO

CO

: "Jj

c

Q

d

n

a

CM

!

Q

Q
UJ

I

c

Q
C-i

T-
1
1

—

u

d

" j

i

a

u

2

•TO
i
i

O
i_j

r,i

1
1

X

X

I

a

CM
i

X

o

t

1

■o

:

j

;i-i

: IL : ■■*-
:iii:G

u_:

T—

W CO

«t

mj(D

N

m

CJ)

o

r-

CD

X

ICO

D

a

d

G

n

i
i

Q

-*

c

cd

i
i

'_ '
a

— '

1

D

c

p-
i

a

C'i

r*
1

D

X

p-
I

X
CO

i

X

x;
dI

l|
j

\*
i n

! _■

! i<

1 ; Z

c5

;0

■cc-uj

5

:U.i£
EC Z

Q
LU

(M 00

h-;i-

rr- ,-r

0:0

Q.:LL
LU:UJ

T
LL

!Q
CL

CC

inito
» — 1 •—

ltiq:

-

cl:u.

UJ.UJ

N

cc

Q
CL

;UJ

ct

LU
LL

CO
h-

a:

Q

LL

UJ

■x

o

LL

hi

N

Q

■0'
CO

" 1 <

i(-
_:oo

— S ; ~r

— l '■ \J

rr\

LUiX
"> : /"

b;T

jure

3

Q
O

d

3

Q

■—

!

o

'."J

I

:
S

7

1

id

liO

1

■

c

ILf

I

o

—

1
)e<

o

j.'
I

:r -

i
1
1

i

1

1
J

_' t

1

1
1

A L
u ■

1

\ o

<

< 1

I o

I

111

Q.

1

n !

i

U !
1

,. .J

I \

I 1
I I

n i

t ■

4- CD

Ml

I I :

—

<^

V.

il-

£.

h

(-

y

CO

; — 'i

HI

J^

!-

u

uu

<_

1

CO

T—

<l

i

u

I I

1, !

-.

n

tir

—

T

V

LU

1

_!

O

I

— •

~7

i

<

V.

X

ii

i—

o

L

,_2

4- in

T M

u

f/T

n

4- co

-t" T-

a

r-j

Lf)

CO
n

e 3.G - St tadii; Ini . ■; Esti.T.aie

P,i o e ', 6

m

m

3 J

MIT

I*!

: ; xvv]

■ififc

p

x&x

is

. ; :«G:
:#•

■to
■■\$■

I-

:>|
>§fl

xWi

xsst

::x'j

I

I
i

■;
i

:

!<B

N

~J

u

i

ioi

(fl:

-CO

LU

EC

•Hi

h-

U

;uj;

i_

:<:

Q

a.

D

;«:

■CM:

in

£0

N

'X'

in:

co-

in

m

a>;

■Wj

:Q:

CM

:_J:

D
uJ

SCMI

:co;

:N

h-
rrv

2

UJ

t_l l_l

o

CO

■I

uJ

CJ

CO

LU

Ui

oia

LU

o

Ml

°l

•— 'j
CO I

j

-

ci
oil

F

INDEPENDENT RESEARCH PROJECT

°.R nn

qn nn

DOUBLE SPiKE iN THE ECAC OF EQUAL MAGNITUDE

n

oc nn _L / a

i \ n

20.00 -a n ft 's 1 i "^

T

5.00 •+

0.00

Figure 3.7 - Double Spike

? a § e 3 5

COSTS i

I j

ic nn

u.uu -j-

41. or'Ar*

i

4 5 6 7 8 9 10

MONTH

COST TREND WORKSHEET

^OU

COS7TZ
BLESPfKi

i m wot

WSHEET

>m&QF THE SAME MAGNITUDE

• i

|

i

■•■!

■ i

ft *TC

23-Mav-90

ORIGINAL COST=

20

1
1

1

i

2

^

4

5

1

I

! PREVIOUS

PROJECTED

REMAINING

ESTIMATED

1

ncruni

nrnnriT

pcrun.i

nrnnnr

: ncruni

i ir-.r- 1 » tt

uruMi c

rnkini r— n.—th.i

Tlk AT-

1 II'IC

UUJ 1 .^ 1

NUMBER

DATE

DATE

INTERVAL i

DATE

TO COMPLETE !

COMPLETION

!

t - i r-, i rr

niru i

INPUT

1

ECAC

atNAL

10

20.00

|

0RT1

1

1

10

9

20.00

,nnT n

Uni -

■ i.

i

i

iu

20.00

;

;

Um d

-■>

1

10

7

25.00

!

n-.nr .(

UP.l *t

4

i

■4 ft

1 u

b

•-.ft nn
iU.UU

1

1

'CRT 5

5

4

1

j :

10

5

20.00

C

C
J

\

4 r.
IU

*t

^!u uu

1
j

|RT 7

7

s

i

10

j

25.00

ifil-iT .-i

■j

7

I ■

Iu

i ■

2u. : Ju

1

1

: 'ORT 9

3

3

1

10

1

20.00

1

'Uru iu

10

1

•4 r,
1 U

-

20.00

i

: : : : • I

J

3

! U

i :

1 J

ESTIMATED

'wwO 1

PREDICTED

PREDICTED

;

LUO i

i-i .— h-vi-

jLurc

h r*. r-. m .—. k i * t
nUUI 1 lUINrtL.

i— -,— i-*T r.T

i^LJ-JI .".1

rirrvcirvr

rcruni

1

OVERRUN

OVERRUN

COMPLETION

NUMBER

I

S~ L'Kiu LLib i

*%t*\..1\

ph ik 4 ft

3/1 2

c. . 4 r,
JA 1 U

8+1 1

1

PCAC

: | j : J

0.00

0.00

0.00

0.00

0.00

20.00

REPORT 1

i

u.liu

u.bo

': U.UU

u.uu

u.uu

2 U.UU

f-.r-r-,/-,r,T -.

ncruni i

I

i

5 00

10.00

10.00

1 11

"7 Ti-i

f . f

32.73

REPORT 3

1

u.uu

u.liu

4 r. r.ft

IU.UU

u.b'o

o.; j

iO.iJ

r. r* n .— t i-it *

ncruni *t

i

0.00

0.00

10.00

0.40

2.00

22.00

REPORT 5

j

u uu

L'.'u'u

1 u.uu

u'.-d

i.ii j

*,4 4 4

nmnnT .*•

ncruni o

1

1

5.00

10 00

20.00

0.41

■* ■■ •"•

26.22

REPORT 7

1

U.OU

.** r,r,
U.UU

>\r, ft ft
..U.UU

ft --.4

U - j 1

U .

^'u'.b'j

n r* n .— . r.T .*.

ncruni o

1
1

0.00

0.00

20.00

0.25

0.25

20.25

REPORT 9

j

u.uu

J 'JiJ

iu.Ou

ft ■-..-.
U.iU

u.uu

2 ! J 00

REPORT 10

1

Figure 3.7 - joubie Spiki

P-iop 33

INDEPENDENT RESEARCH PROJECT

DOUBLE SPIKE iN THE ECAC. ONE INCREASE AND ONE DECREASE

45.00 -r g

4n nn -i- / \

w nn

25.00 4-

!

COSTS !

I

20.00 I

0.00

/

30.00 4 I \

\

I

! Lk

;

(

/ /

i

l-

~~~----r\

lt- — _n

— i 1 r

1 1

i 1

1 !

i_i

u

u

1

L_

15.00 4 C

I
!

10.00 4-

i
i

- „- i

b.OU -+■

4 5 6 7 3 9 10

MONTH

Figure 1 . : , - Increase :pii:e - i ' i. .
Folluwinu Decrease ^pike

Pa«e 4

nnsT Tr.FNn wrtRKSHFFT

DOUBLE SPtKE IN THE ECAC ONE INCREASE AND ONE DECREASE

URIGINAL COST=

20

rerun, i

PREVIOUS

PROJECTED

REMAINING

uruni c

EST MATED

•— - .— ■ .---r .'• t

LUJ I .1 I

NUMBER

DATE

O.-.TE

INTERVAL

DATE

TO COMPLETE

COMPLETION

COL4 -CQL1

UNA

PRT1

10

fcUAL
20.00

10

:i!.UU

IQRT3

10

iJ.UU

20.00

10

2U.UU

OPT 7

10

15.00

lU.'JU

;'UPT 3

i n

2U.UU

ESTIMATED

OVERRUN

CuST

PREDICTED

OVERRUN

PREDICTED

LUOI .". I

COMPLETION

NUMBER

S- ORIG COST

2!3x

0.00

0.00

0.00

0.00

-'UAL

IJ 00

:0.00

REPORT 1

J.UU

0.00

Q.QO

0.00

0.00

10.00

REPOP;

0.00

10 00

0.63

u uu

10 00

0.40

z.uu

0.00

1 o.oQ

22.U0

REPORT 4
.REPORjT

r-. r- r. .— i-.t .-■

-iu. oo

o.oo

0.00

0.00

ou

REPORT 7

o.uo

o.oo

o.ou

00

o.oo

o oo

20.00

riruni o

"reportT

Figure 3 . S - Increase Spik" witii .

Foil iwi;i? Dei raase Spike

Pat>e H

3.4 SENSITIVITY OF THE WORKSHEET

The formula used in the worksheet is described in subpart
3.2. The two most easily manipulated factors of the worksheet,
without changing the integrity of the worksheet, are found in
columns 8 and 10. Both of the factors are described below.

COLUMN 8: Column 8 is a calculation column, it multiplies
the product of the update interval by the estimated cost
overrun by a factor of 2. This factor of 2 is interesting
in that it increases the cost overrun which is later
compounded and then averaged to determine the predicted cost
at completion, PCAC. If this factor is reduced, the PCAC is
reduced. This can be seen by reviewing Figure 3.8 -
Sensitivity Analysis, Column 8 Factor Set at 1.0. By
reducing the factor in Column 8, from 2 to 1 the PCAC is
reduced. This is shown by comparing the Altered PCAC Column
8 factor of 1 to the original PCAC, Column 8 factor of 2.
When the factor was increased, from 2 to 3 , the PCAC
increased, Figure 3.9 - Sensitivity Analysis, Column 8
Factor Set at 3.0. This is shown graphically by comparing
the Altered PCAC, Column 8 factor set at 3 and the original
PCAC, Column 8 factor set at 2.0.

42

o

i

<

o

Q_

Q

LLi

■T

o

lL

<

h-

<

o

HI

<

CL

■

n

4

i_i

' 1

1

V- 1 t-

X

i

j
t

1
n
i
ui

A

X m

OJ

LL

5£

CO S=
LLi H

■U"\

\ \ \

">dra

/ 1 i
/ 1 1

/ / 1

i * a

/ / i

/

i
/ i

i - -

T

■_

LS

***

T

o
d

■—■

i —

'_

c

Q

3
CO

CM

CO
u

CM

Lfl

F Lgure

3.9 -

Sensit i

. i t \

kna lysi

Column -

:".i. ;

or

Set

at

1.0

\a

?a«e 4 3

COST TRPNP WflWCRHPPT

COSTTRENO: '-'WDPKSMEEX

SENSITIVITY ANALYtSlS - COLUMN: ft F ACTOR SET AT**

fi * TT- AC ».#-•• AA

ORIGINAL COST= : 20

12 3 4 5 S

: PREVIOUS: PROJECTED REMAINING ESTIMATED

nmnnr nrnnnT nrnnnT i mn*Tr r--r-.kj(rn r—rifiki tuj^ r^nr^T at

ncruhi ntruni ncrurvi . uruMtc uurtruc I iUiv uric luoi .-. i

NUMBER DATE DATE ! INTERVAL : DATE ! TO COMPLETE ; COMPLETION

ikini it r-r-.i a rni 4 >MC*I | " r

iNrul luli'luLi iftrUi

ECAC

ilNAL 10 20.00

0RT1 1 1 10 3 20.00

i~.nT a a 4 4 4 a O ^^ ^f>

)_;n.i i iii rj ij.uu

OPT 3 3 2 1 10 7 20.00

ir-irvr 1 * a 4 4 n a ^n aa

uni *t *t J i iu o ..u.uu

OPT 5 5 4 1 10 5 20.00

orvr a r> c 4 4 a 4 r .c nn
L'HI 0. J 1 IU t iJ.UU

0RT7 7 6 1 10 3 20.00

r~> fat n a -? 4 4 a a a a a a

um oil : u l *u.uu

OPT 3 3 3 1 10 1 20.00

j— .FIT 4rt 4 A A 4 4 A A Aff AA

uni 1 *j iu : 1 lu u iU.JU

;

■7 A A 4 A 44 4 A

1 j IU II I £

ESTIMATED COST PREDICTED PREDICTED

r^r-.r-T pi r-M-vr- »r,niTinkui r--r-.<--T /■ T r,r-r-.«— -tat

wUJi OLurc rtuui 1 rjHrtL luoi .ni rcr uru

OVERRUN OVERRUN i COMPLETION NUMBER

» r->oir-> rnrr a*a.."v. n ih 4 a GM''** E«. in A .4 4

D-urUbuuol ilOArJ oui'i o d/l i w iu o+l i

PCAC

! ! ■ ' '.

■■ i

0.00 00 0.00 0.00 0.00 20.00 REPORT 1

C AA 4 A AA 4 A AA A CA AA AA »C AA FA f f* PA TAT A

u.uu 1 u.uu 1 u.'j'j i.ou iu.uu nu.uu n.crum £

0.00 0.00 10.00 1.11 7.73 27 73 REPORT 3

A aa A aa 4 a aa a aa a -?p aa ->e nrnnnT 4

u.uu : u.uu mJ.Uu J. 00 j.i j xo.ru nrruni n

0.G0 0.00 10.00 0.40 2.00 22.00 REPORTS

C AA 4 A TiA AA AA A CA A AA A 1 ^ AA ft F~ TA PA FAT A

j.uu iu.uu „u.uU u.uo i.zz ^r .ii her Uni o

0.00 0.00 20.00 0.41 1.22 21.22 REPORT 7

A AA A AA AA AA A A4 A AA AA AA ArnnPiT A

u.'j'j u.uu iu.uu u.oi u.oo iu.oo ri^r'jPvi o

0,00 0.00 20.00 0.25 0.25 20.25 REPORTS

A AA A AA AA AA A AA A AA AA AA nmAHT 4 A

U.UU u.uu iU.UU U.i.u u.uu *.u.uu r-cruru iu

figure 3.9 - Sensitivity \\
"olumn S Factor 5^; nt

ilvsis,

'atie i-t

nnsT TRPwn wdrushfet

•:•:•:•:•:•:•:•:•:•*:•:•:•:•:•••:•

:•:•:•:•:•:•:•:•:•:•'■ • •■■•'•'•••••'••• :•:• • :•:•

ESTIMATED

COST

PREDICTED

PREDICTED

COST

SLOPE

COST AT

RfcPORl

OVERRUN

OVERRUN

COMPLETION

Kll IMP.CP.

.WW. .

- ORiG COST

(3x7)

SUM 8

*/n

5x10

6+11

ALTERED-PCAC

Q 00

00

■ 0.00

0.00

00

20 00

REPORT 1

5.00

5.00

5.00

1.25

10.00

35.00

REPORT 2

0.00

0.00

5.00

0.56

3.39

23 S3

BtrpfiRT ^

0.00

0.00

5.00

0.31

1 .00

21.88

REPORT 4

i'i Qlj

I Q yi]

5.00

20

1 .00

21 .00

REPORT 5

5.00

5.00

10.00

0.28

1.11

26.11

REPORT 6

0.00

0.00

10.00

0.20

0.61

20.61

REPORT 7

0.00

0.00

10.00

0.1b

0.31

20.31

REPORT 6

00

0.00

1 00

0.12

0.12

20.12

PEPORT 9

0.00

0.00

10.00

0.1

0.00

20.00

REPORT 1

figure 3 . 3
Co] umn

Sensi i ^ v 1 1 \ Ana . ls,
ict or Set at i .

"age -ij

INDEPENDENT P.E2EAP.CH PQGJECT

60.00 j
50.00 -

40.00 --

!

I

^otq on no.

SENSITIVITY ANALYSIS

//
//

n
u

.1 \

/ \

/'♦ \
/ A \

\V

COLUMN 8 FACTOR SET AT 3.0

9 --" v -^ ^U ^-KjTvi-'''S*.."V _

20.00 s — jjp- ^ — a — ^r ^s=Sh-#

1- ECAC

J~L ii -nrncrr* nfJi
i_T «L i L-ru-i^-rOM'

-A- OpAp

10.00 4-

0.00 4-

1

4 5 6 7
MONTH

10 11

Figure 3.10 - Sensitivity Anal-. i<
Ci i i u:nii 3 Fai tor Set it 3.0

16

COST TREND WORKSHEET

" VCOST.TREMT l^Mt

'■'.":'.■■■■'■'■"■"■''■

P"> (.TT

25-May-90

ORIGINAL COST= i

20

i

A

a

4

c

A

PREVIOUS

PROJECTED

REMAINING

ESTIMATED

REPuRl

ntruni

rtr-r-ii—tr-t-r

nirjRi

i inn 4TP

uruni c

nnuni rnnM
'-UI'iri_C HUH

Tikir
i iric

r-nrT &T

NUMBER

date

DATE

INTERVAL !

DATE

TO COMPLETE

COMPLETION

ikir-.i rr

iwrul

'wULI *LULI

ikir-.i it

ECAC

ilNAL

10

20.00

3RT1

1

1

10

3

20.00

Sot n

a

i

1 :

4 .-■

A

AC AA

iJ.UU

DRT3

i

•>

1

10

7

20.00

"ArVT 1

_ini t

4

3

1

10

A

A A A A

£U.UU

DRT5

5

4

1

10

5

20.00

JP.I

a

e
j

]

4 A

i u

t

AC AA

DRT7

7

i :

b

1

10

20.00

".RT .".

jni o

a

i

1

4 a

i u

A

A A A A
-U.UU

DRT3

3

3

1

10

1

20.00

nr.T -4 a
J hi iu

4 A

1 u

3

4
1

4 A

A
'J

20.00

1

i :

a
O

A

10

4 4
1 1

1 £

ESTIMATED

COST

PREDICTED

PREDICTED

nnrT
LUO 1

pm nnp

dLurc

* nnmnk i a i
rtUUI 1 l'JI1.".L

PT-.r^T AT

LUJI .-.1

r.r- ^-l^">r■^T
Kc^'unl

OVERRUN

OVERRUN

COMPLETION

NUMBER

S .— .nir"> nr.rT

3- uruu luj 1

A/ A. .-?>
4oX( ]

CM Ik J A

OUI'1

6f\ L

5x10

6+11

PCAC

0.00

0.00

00

0.00

0.00

20.00

REPORT 1

c aa
J.UU

IU.UU

4 a aa
IU.UU

A CA
...w'U

aa rtft
iU.UU

iC A A

nu.Uu

r.rnppiT a

xruni ^

0.00

0.00

10 00

1.11

-J 70

f Y o

27.7S

REPORT 3

0.00

A r>r,

u.bu

1000

0.83

O.f J

AA "?C

REPORT 4

0.00

0.00

10.00

0.40

2.00

22.00

REPORT 5

j.Oli

10.00

20.00

0.58

? **0

•".-» A A
♦if .LL

p-i p" n (—< DT C

ncruni o

0.00

0.00

20.00

0.41

1.22

t 1 ,Xx

REPORT 7

n a a

U.'JU

a ap.

U.UU :

A A A A

.*■ A 4

U.O 1

p. a a
V 00

A A PA

iU.OO

nrnnnT a

ncruni o

0.00

0.00

20.00

0.25 i

0.25

20.25

REPORT 3

0.00

0.00

a a r,r,

iU.UU

0.20

A A A

u.uu

20 00

REPORT 10

Figure 3.10
Co lumn

Sen .. . t ; ■• i t v Ana lysis ,
Factor S^t at 3.0

Pase J'

mSTTRFWn WnRKRHFFT

1

::::::::::*:::::::::::x

ESTIMATED COST PREDICTED PREDICTED

COST AT

REPORT

OVERRUN OVERRUN

COMPLETION

NUMBER

rORIGCOST \3»{3x7) SUM 3 3/1 "2 5x10

6+11

ALTERED-PCAC

0.00 0.00 0.00 0.00 0.00

20.00

REPORT 1

5.00 15.00 15.00 3.75 30.00

55. 0Q

REPORT 2

0.00 0.00 15.00 1.87 11.67

j1 b7

REPORT 3

0.00 0.00 15 00 0.34 5 93

25.63

REPORT 4

0.00 £#0 15.00 0.60 3

00

23.00

REPORT 5

5.00 /5100 30.00 0.83 3

33

28.33

REPORT 8

0.00 0.00 30.00 0.61 1

34

21.34

REPORT 7

Q.OQ 0.00 3000 0.47

94

20.34

REPORT 3

0.00 0.00 30.00 0.37

37

20.37

REPORT 3

0.00 0.00 30.00 0.30

00

20.00

REPORT 10

Figure 3.10 - Sensitivity Analysis,
Column S Fictor Set it 3.0

Taw,. 15

COLUMN 10: Column 10 calculates the cost slope by dividing
the sum of the estimated cost overruns by the report date
squared. The factor to be considered in this column is the
square of the report date. If the report date is raised to
a power greater than two, the PCAC will be reduced further.
If the report date is raised to a power less than two the
PCAC will increase. Figures 3.11 and 3.12 show the change
in the PCAC when the report date is raised to the power of 4
and when the report date factor is not raised to any factor.
Figure 3.13 - Sensitivity Analysis - Cost Slope Factor
Analysis shows a one time increase in the ECAC with
consistent ECAC's following. This figure shows that with a
Cost Slope factor of 1, the profile of the Altered PCAC arcs
shortly after the increase in the ECAC therefore continuing
the amplification scare factor for 3 additional reports
after the ECAC has increased.

49

INDEPENDENT RESEARCH PROJECT

SENSITIVITY ANALYSIS

.•US I e>

45.00 T

!

40.00 4

I

35.00 --

30.00 --

I
25.00 J.

I
20.00 #

15.00 -

iu.uO -f

5.00 I

0.00 4-

1

COST SLOPE FACTOR CHANGED TO THE POWER OF 4

K
\

!

\

A

\

/

i

A
•-'n\

/*

/ / ■ \

4 5 6 7
MONTH

10 11

+ ECAC

-Lr ALTERED-PCAC
+ PCAC

Figure 3.11 - Sensilivity Analysis,

:ost Sloj.e Factor Changed

to tiip Power of 4

Page 50

COST TREND WORKSHEET

:COSTtreno;: ^worksheet;.

ifcaa^^

r-» 4TT- ac ki... aa

ORIGINAL COST= j 20

12 3 4

C A
J

! PREVIOUS ; PROJECTED

REMAINING ESTIMATED

rvr~rtnr%T nrnnnr nrnnPiT i innaTr r-nkini rnnn

nfcruhi rtruni rvcrurvi : uru. u .ic uurirLciiuii

-m if rnrvr at
iiimc uuoi r. i

NUMBER DATE DATE | INTERVAL | DATE

TO COMPLETE i COMPLETION

INPUT

C0L4 -C0L1 INPUT

ECAC

3INAL 10

20.00

0RT1 10 1 10

3 20.00

prvr a 4 4 i a
uni £ £ 1:1 i u

A ".C A A

^C'.uu

0RT3 3 2 1 10

7 20.00

f*rvr 4 4 a 4 4 o

sni t 4 o 1 id

b iU.uu

0RT5 5 4 1 10

5 20.00

.nrvr a .a c * « ft

uni o o j i iu

4 AC AA

4 ij.uu

0RT7 7 6 1 10

3 20.00

r\rrr ft a i ^ * ■-,

uru o o i l iu

-i A A A A

£ i'J.UU

0RT3 9 3 1 10

1 20.00

nnr ha * a a * * a

uni iu iu o i iu

A A A A A
U 4.U.UU

*> A A 4 A « 4

i J 1 U II

12

ESTIMATED i COST PREDICTED

PREDICTED

n>ri£T rn nnr * r-ir">mr-ikt * i
UUOI JLUrC rtUUI 1 HJI1ML.

,-*■ r-, „- t at r-. r- r-. .— . i-it

uuoi m nfcruni

OVERRUN OVERRUN

COMPLETION NUMBER

fi_ rvQif^ rnCT : A*A..-n r-i ik 4 r. a>4 a a c. . 4 a

ur.ia luo i iioxi j auri 0/1 i ua iu

6+11

PCAC

0.00 0.00 0.00 0.00 0.00

20.00 REPORT 1

P AA 4 A AA 4 A AA A CA AA AA

J.UU 1 U.UU IU.UU : L.JV : iU.UU

4C AA ni-r-.r". r-.T -.

iu.uu Kcruni l

0.00 0.00 10.00 1.11 7.73

27.73 REPORT 3

0.00 0.00 10.00 0.83 3.75

23.75 REPORT 4

0.00 0.00 1000 040 2.00

22.00 REPORTS

5.00 10.00 20.00 0.56 2.22

27.22 REPORTS

0.00 00 20.00 0.41 1.22

21 .22 REPORT 7

f\ ftfl ; A AA AA AA A A4 A AA

u.uu u.uu id uu u.ji u.oo

ft ft r»n r-.r-rw— .rvr n

iu.oo ncruni s

0.00 0.00 20.00 0.25 0.25

20.25 REPORTS

000 0.00 20.00 0.20 0.00

20.00 REPORT 10

Figure 3.11 - Sensitivity Analysis,

Cost Slope Factor Changed

to t, lu j Power of !

Page 51

COST TREND WORKSHEET

ESTIMATED : COST PREDICTED PREDICTED

COST AT REPORT

OVERRUN OVERRUN

COMPLETION NUMBER

-ORIGCOST 2*f3x7l SUM 9 \ 9/1*4 5x10

6+11

ALTEREO-PCAC

0.00 00 0.00 0.00 0.00

20.00 REPORT 1

5.00 10.00 10.00 0.S3 5.00

30.00 REPORT 2

0.00 0.00 10.00 0.12 0.88

20.86 REPORT 3

0.00 0.00 10.00 0.04 0.23

20.23 REPORT 4

0.00 0.00 10.00 0.02 0.08

20.08 REPORTS

5.00 10.00 20.00 0.02 0.06

25.06 REPORTS

0.00 0.00 20.00 0.01 0.02

20.02 REPORT 7

n nn qoq 20.00 O.QQ 0.01

20.01 REPORTS

0.00 0.00 20.00 0.00 0.00

20.00 REPORT 9

n nn n nn >n nn nn n nn

W.WW W.WW ^W.wtj •#• •-# u w.ww

'/n nn pcpnoT 1 n

Figure 3.11 - Sensitivity Analysi .,

Cost Slope "actor Changed

to i he Power of 4

Page 52

INDEPENDENT RESEARCH PROJECT

SENSITIVITY ANALYSIS

COST SLOPE FACTOR CHANGED TO THE POWER OF 1

-a- ECAC

Figure 3.12 - Sensitivity Anal-, .
"st Slope Factor Changed
!' f lie Power >f 1
Page "

COST TREND WORKSHEET

cost mem womsmer

SEHSmVlW ANAW51S -COST SUJPE I^TOftCHiVMSEO TO THE POWER OP t

DATE 25-May-90

ORIGINAL COST= \ 20

■* ^ A J

C A

"J O

1 PREVIOUS ! PROJECTED

REMAINING ESTIMATED

T~,t-nr\r~»-r ni-r%o'~>T nrnnnT i irt-in*Tr~ rnkjni r— pr-.ki

nfcrurxi n£runi rtrunl : urbnic UurlrLc.UUn

uric Luoi r,\

NUMBER DATE DATE i INTERVAL ; DATE

TO COMPLETE j COMPLETION

ikini rr

mrui

CCL4 -CCL1 INPUT

ECAC

3INAL 10

20.00

0RT1 10 1 10

9 20.00

DRT 2 2 1 1 10

A AC AA

iJ.UU

0RT3 3 2 1 10

7 20.00

0RT4 4 3 1 10

6 20.00

0RT5 5 4 1 10

5 20.00

nr.T r- a c < < a

uni o o u i Iu

4 25.00

0RT7 7 6 1 10

3 20.00

nrvr a a -? « m

un io o r i i u

A AA AA

i i.U.UU

DRT 3 9 8 1 10

1 20.00

HOT HA 4 A A 4 4 A

_*nl iu iu d i iu

A AA A A

u iU.UU

"» A A HA 4 4

( 1 U It

1 L

ESTIMATED COST PPEDICTED

PREDICTED

r^r-if-T at r-.r-r^r-ir^T

luji .-,i ncruni

OVERRUN OVERRUN

COMPLETION NUMBER

6-ORIGCCST 2(3x7] SUMS 3/1 "2 5x10

6+11

PCAC

0.00 0.00 0.00 00 0.00

20.00 REPORT 1

J.uu ib.uu lu.uu i.ju iu.uu

iC AA r-V^f-if-iflT A

nu.uu rcruni l

0.00 0.00 10.00 1.11 7.73

27.73 REPORT 3

A .*.A : A AA -1 A AA A A A A "?C

U.'JU U.UU lU.UU U.OO O.IU

aa fc nrnppT .4

000 0.00 10.00 0.40 2.00

22.00 REPORT 5 1

5.00 10.00 20.00 0.56 2.22

A-» AA J-iT-r-tr-.l-iT A

ii .ll rcruni o

00 0.00 20.00 0.41 1.22

21 .22 REPORT 7

A AA A AA AA AA A A4 A AA

U.UU U.UU ilU.UU U.Jt u.oo

AA AA P", p- TA rs r.T A

*.u.oo nr.ruai o

0.00 0.00 20.00 0.25 0.25

20.25 REPORT 9

0.00 0.00 20.00 0.20 0.00

20.00 REPORT 10

Figure 3.12 - Sensitivity Analysis,
Cos1 Slope Factor Changed

t i) r he Power or 1
Pa me i>4

rnsTTRFhin WORKSHEET

7 v

.Y.Y.Y. Y.TY. *.Y. Y.Y.Y .Y.Y.Y.

ESTIMATED

y"yX\y"?!y. y.y.y.

WM&M

mm

COST

Iff ." '.

PRtUiCTED

12

PRhUiCTEU

: -: : : : x : >: : : : >: : : : : : :y: : : : : : : : :y': : : : : : : : :

COST

SLOPE

COST AT

REPORT

0VER8.UN

OVERRUN

COMPLETION

NUMBER

6- ORIG COST

2*(3x7)

SUM 3

'9/1

5x10

6+11

ALTERED-PCAC :

i i

0.00

0.00

0.00

0.09

0.00

20.00

hEPORT 1

5.00

10.00

10.00

5.99

40.00

65.00

REPORT 2

0.00

0.00

10.00

3.33

l-i.Si

4d.oo

REPORT 3

0.00

0.00

10.00

2.59

15.00

35.00

REPORT 4

0.00

0.00

10.00

2.99

10.00

30.00

REPORT 5

5.00

10.00

20.00

3.33

13.33

38.33

REPORT 6

0.00

0.00

20.00

2.36

8.57

28.57

REPORT 7

00

00

20.00

2.59

5 00

25.00

nCPHPT 5 i

0.00

0.00

20.00

2.22

2.22

22.22

REPORT 3

00

00

20.00

2.99

0.00

20 00

REPORT 1

Figure 3.12 - Sensitivity uial . »is,

; .•■ Hope Fa to ( hanged

to ' he Power of 1

INDEPENDENT RESEARCH PROJECT

p.n nn _
I

I
7n nn -i

i

60.00 +

50.00 |

I

T

I

30.00 +

20.00 I-

i

I

10.00 +

^r»IAITIUITV/ A It A I W/NI/*

Oil!Mv>! ! IVM T MfMMLTOiO

r*r\c*-r oi one c»o--iv-iri »n*i voio

„-U — -TL

rv--
U

Uv.

Lk.

/ f ^

/ / . ^---

/ / (^

Ik

if/ ^ I-

I -I- ECAC

i H> ALT A 1-Pf:AC:

I

I

i + PCAC

\ I

^C I -0- ALT A 3-PCAC

0.00 4-

-i 1 h

4 5 6
MONTH

10

Figure 3.13 - Sensitivity Anal >i .
Cost Slope Factor -Uialvsis

TOST TREND WORKSHEET

COST TREND WORKSHEET

DATE 27-May-30

ORIGINAL COST= j 20

•* A A A P A

1 L 1 1

; PREVIOUS! PROJECTED REMAINING ESTIMATED

nmnrvr rAi-p>r-.r.-r r-.r-r-.r-.r-iT i u-,r"\*TT- r--rik jr-.i r — nr.ki -tuji- rnrT at

niruni ncruni ncruni : urunic uurirLC nun uric uu^i rti

NUMBER DATE DATE ; INTERVAL j DATE j TO COMPLETE j COMPLETION

ikir-.i rr t--r\\ j rni 4 ikir-ii rr

irirul uuLi-Luli irtrul

ECAC

3NAL 10 20.00

QRT1 10 1 10 3 20.00

nnT a a ■* ■* ^ >*> **' *%c Tfrt

uni ^. ^.ii i u o ij.uu

0RT3 3 2 1 10 7 25.00

r-ii-.x j j a ■< < n a ac .-.a
uni n f o i iu o iJ.uU

0RT5 5 4 1 10 5 25.00

nrYT "■ ^ c 4 ■* a j nc a a
UP.l J 1 1 U *t iJ.UU

0RT7 7 6 1 10 3 25.00

f.rVT n A -J 4 4 A A AC AA

uni o o t t i u ^ ij.uu

terra 3 3 1 10 1 25.00

nrrr *n * a a 4 4 a a ac r.r.

un.i iu iu j i iu u ij.uu

■? A A 4 A 44 4 A
1 1 U 1 1 \ L

ESTIMATED COST PREDICTED j PREDICTED

r-n^T r-i r^rAr- h nr.mr.k > 6 i r^r-vr'-r a -r n r- ia n, r.-r

uuai jLurc MUUihur4.nL uuoi «i ncruni

OVERRUN OVERRUN COMPLETION NUMBER

a -.r-,ir^ r-'.—.r^T a/a..->'» nt iu a a.'4 A a c. 4a a, 4 4

o uniu luo i uo/m I ouM o d,' i i ja iu o+ 1 I

PCAC

!!': ! 1

0.00 0.00 0.00 0.00 0.00 20 00 REPORT 1 I

C Art 4 A AA 4 A AA A CA AA AA (C AA nmnAT A

j.uu iu.uu iu.uu *..ju -u.uu nj.uu rcruni i.

5.00 10.00 20 00 2.22 15.56 40.56 REPORT 3

c aa 4 a aa in aa 4 aa 4 4 ic ip ic npnnnr .* i

j.uu iu.uu ou.uu 1.00 11. i3 00. i. j KfcrUfti 4

5.00 10.00 40.00 1.60 3.00 33 00 REPORTS

P Art 4 A AA CA AA 4 AA C CA AA CA T~, r~ ft f~ < AT A

3.UU i'j.uu ju.uu i .od D.jd ou.uo rtruni o j

5.00 10.00 60.00 1.22 3.67 23.67 REPORT 7

C Art -1 A AA "?rt Art 4 rtA A 4 rt AT 4 A nrnnriT A

-j.'ju iu.uu iu.uu i Mo i.io ir.io rcruni o

5.00 10.00 30.00 0.33 0.33 25.33 REPORT 3

C AA 4 A AA AiA AA A AA A AA ^C AA P"i P" O r"> HiT 4 A

■j.u'j Iu.uu ju.ju u.ju u.uu £D.UU her uni iu i

Figure 3.13 - Sensitivity Analysis,
Cost Slope Facior Analysis

rnsTTRFNn ■jdmsmfft

yXy.y.y.t.y.y.y.y.y.y.y.y.

ESTIMATED

xxyX-x*X£x-xx

.Y.V.Y.V.Y.V.W.Y.

: : £u\$T SiuPfc FACTtJH S£T AT f v y-
W/M-i ■ '■ •••+*•• ••• ••'••'"•• • -»+ ••

\Y \'XyX'X\y/Xy * ' * ' •'•"•'•'•'.y.'.t.y.y.y.y.y.y.y.y.y.y.y. '■ r.Y.Y.Y. y.y.y

i COST PREDICTED

x : : : x : : : : ; : : : : :x : :+*: : : : : : : : : : :v: : : : : ; :: : .

v.v.v.v/.v.v.tjv.*. .y.y.v.v.v.

PREDiCTED

i

••::.:.:v-.::.::.::V:::..|
■xvxvx::vxxx:vx:x---x|

I

COST

SLOPE

COST AT

pcpnPT

nc. _. •.. |

OVERRUN

OVERRUN

COMPLETION

NUMBER

r QPJG COST

2*f3x?;i

SUM 3

'9/1

5x 1

6+11

alTi-pcac

i

: : I : : 1
: : : : : 1

0.00

0.00

0.00

O GO

0.00

20.00

REPORT 1

5.00

10.00

1 00

5.00

4(1.00

65 00

REPORT 2

5.00

10.00

20.00

S.S7

46.67

71.67

PEPORT 3

5.00

10.00

30.00

7.50

45.00

70.00

REPORT 4

5.00

10.00

40.00

3.00

40.00

65.00

PEPORT 5

5.00

10.00

50.00

8.33

33.33

58.33

REPORT 6

5.00

10.00

60 00

8.57

25.71

50.71

REPORT 7

5.00

10.00

70.00

8.75

17.50

42.50

PEPORT 3

5.00

10.00

80.00

3.39

8.83

33.39

REPORT 3

5 00

10.00

90.00

9.00

.00

25.00

REPORT 10

ffiST <ii fl

m FACTOR

eprcraa/ff-sx-Xv

. . uu-si jcl

•:•:•:•:•:■:•:•:•:•:•:■):+:•:::•:•:■:■:•:■:•:•:•
PREDiCTED

t'/
PRtDiC ! tU

i

ESTIMATED

COST

COST

SLOPE

AOOmONAL

COST AT

PC PHOT

. ,_. w. .. j

OVERRUN

OVERRUN

! COMPLETION

NUMBER

r OPJG COST

2**3x7)

SUM 3

'9/1*3

5x 10

6*11

i

AL1 3-PCAC

: • ! :

• : : : i

0.00

0.00

0.00

0.00

0.00

20.00

REPORT 1

5.00

10.00

1 00

1.25

10 00

35.00

REPORT 2

5.00

10.00

20.00

0.74

5.19

30.13

PEPORT 3

5 00

10.00

30.00

0.47

2 31

27.81

REPORT 4

5.00

10 00

40.00

0.32

1.60

26 60

REPORT 5

5 00

10.00

50.00

0.23

0.33

25.93

REPORT 6

5.00

10.00

60.00

9.17

0.52

25.52

REPORT 7

5. 00

10.00

70 00

0.14

0.27

25.27

REPORT 3

5.00

1 00

30.00

0.11

0.11

25.11

REPORT 9

5.00

10.00

90.00

0.09

0.00

25 on

REPORT 1

Figure 3.13 - Sensitivity Anai\ ;i ,
Cost Slope Factor Anal

'a^e

Table 3.1 - Results of the Sensitivity Analysis tabulates
the results of the Sensitivity charts and worksheets. For each
different change in the worksheet a row is provided showing the
percent change in the ECAC, percent change in the PCAC and the
percent change in the Altered PCAC. For each item the percent
change in the ECAC and the PCAC are kept constant therefore a
comparison can be make to the Altered PCAC. The 3 0% and the 7 0%
project completion status is also shown for each change in the
different sensitivity analysis charts.

By reviewing this table, it can be noted that changes at the
30% level of completion can easily be manipulated to change the
amplification factor. For example, be changing the Column 8
factor from 1 to 3 the amplification factor increased from 75 to
175%. The lowest amplification factor is 50% at the 30% complete
time period for the Cost Slope Power equal to 4.0.

SENSITIVITY ANALYSIS RESULTS

CHANGE IN THE FORMULA

% CHANGE
IN THE ECAC

% CHANGE
IN THE PCAC

% CHANGE IN THE
ALTERNATE PCAC

PERCENT COMPLETE

30 %

70%

30%

70%

30%

70%

COLUMN 8, FACTOR - 1.0

25%

25%

125%

23%

75%

24%

COLUMN 8, FACTOR ■ 3.0

25%

25%

125%

23%

175%

23%

COST SLOPE, POWER = 1.0

25%

25%

125%

23%

225%

28%

COST SLOPE, POWER = 4.0

25%

25%

125%

23%

50%

25%

Table 3.1 - Results of the Sensitivity Analysis

59

Several deductions can be made by analyzing the sensitivity
analysis:

1. It is easy to manipulate this worksheet.

2. Davis's intention was to have a large amplification
factor, which would shock the project manager into action.

3. Very small changes at the 70% level compared to changes
at the 30% level. This is called the convergence of the
PCAC to the ECAC. Davis is stating that the estimate
"busts" at the beginning of the project are more critical
that the same estimate "busts" towards the end of the
project.

60

The two problems raised so far in the analysis of the
spreadsheet is that the amplification factor is too large and
that the convergence of the PCAC and the ECAC is not fast enough.
Since we have determined that the worksheet is easily manipulated
by changing the two factors in columns 8 and 10 as discussed
earlier, we can now improve the spreadsheet with the ultimate
goal of reducing the amplification factor and increasing the
convergence .

To reduce the amplification factor the Column 8 factor is
held constant and Column 10 factor is increased to 3 . . It
should be noted that the amplification factor can also be
decreased by simply reducing the Column 8 factor, but this
solution would do little to increase the convergence of the PCAC
to the ECAC. Therefore the ultimate solution that would solve
both of the improvement parameters is to increase the Cost Slope
factor to 3.0. These results can be seen by reviewing Figure
3.14 - Improvement Analysis. This chart shows the difference
between the two options detailed below:

Alternative A: Cost Slope Set at 3.0 and the Column 8 Factor

set at 2.0.

Alternative B: Cost Slope Set at 3 . and the Column 8 Factor

set at 3.0.
It would seem that Alternative B is the optimum solution due to
the large scare factor but with an accelerated convergence.

61

Other factors to consider in improving the worksheet are the
following items:

1. ECAC reduction so large that PCAC becomes negative .
There are several ways in which thus situation can be corrected:

a. Program spreadsheet to decrease the PCAC only by a
fraction of the normal amplification factor. Therefore
have two methods one of increases in the ECAC and a
less drastic method for decreases in the ECAC.

b. A much more simpler method would be to decrease the
amplification factor, but by doing this the original
intent of Davis's spreadsheet is lost.

2 . Begin method at the 20 to 255 work in place stage. In
Chapter 2 several current methods of forecasting were introduced.
In the majority of theses cases the project managers did not
begin using the method until the 20 to 25% work in place stage of
the project. This method should also be used in the same manner,
if used prior to the 25% complete stage the amplification factor
is largest and might be neglected by the project manager.

62

45.00 -r-

40.00

INDEPENDENT RESEARCH PROJECT

i»/innr>v;r nr - kit amai v*"""'^
imr nwvizivirziN « nnnLioio

iORTR

00. UU -j-

on pn i
ju.uu -r

25.00

20.00 i-

// \
i I , \
1 1 * \
// / \ \

i/ \ A s

1/ /

17/ a

n

w

f

i

i

u<

"Ik

Ik.

s 1 a — \$ *=£

"-fL

i -■- tCAC

I ir PCAC

i" ftLi A ■ r>.-rtL

ib.UU

10.00

5.00

Q.QQ

1

4 5 6
MONTH

10

Figure 3.14 - improvement Anal>

Page 63

nnsTTEFNn wdrkshfft

kj*j&f f.ftciw- nrU4
iMPRQVi

4KSffCCf

EMENT ANALYSIS

...Y.Y..Y.Y.Y.....Y.......Y.1

•:y::y: : : : : : :y:y: ; : ; :y:y.y.y: : :y:y]

r-» .. tt-

c 1. ._ r.r,

j-jui rou

ORIGINAL COST=

20

1

ft

4

e
J

r- I

I

1 PREVIOUS

PROJECTED

REMAINING

ESTIMATED

REPORT

i REPORT

: REPORT

UPDATE

COMPLETION

TIME

LUbi AT

NUMBER

DATE

date

INTERVAL ;

DATE

! TO COMPLETE j

COMPLETION

INPUT

CQL4 -COL1

INPUT

ECAC

" Ik l n 1

ft

u

ill

r.r. r.r. 1
iU.UU

0RT1

■1

i

1

10

3

20.00

nnT -

UP.I I

a

4

1

■4

:

4 ft

1 U

-.z r.r.

iJ.UU

0RT3

J

L

1

10

7

25.00

nnr j
'-JP.1 *t

n

A

4

i

4 ft

1 u

ft

-,c r.r,
iJ.'J'J

0RT5

5

4

1

10

5

25.00

r-.r.-r .-■

gni o

r-

J

1

4 ft

; U

t

:z r.r. !
-J.UU

CRT 7

7

s

1

10

3

25.00

Lin. i o

■7

4
1

4 .-.

1 u

: i

" -c'r.A !

i J.U'J

0RT3

9

3

1

10

: 1

25.00

uhi i u

4 ft
1 u

ft

4
1

4 ft

1 u

u :

*CAft !

T

ft

3

4 ft

1 u

4 4
1 I

! ii

j

ESTIMATED

COST

PREDICTED

PREDICTED

:

n^i-iT

'wUO 1

JLurc

uuo 1 A i

REPORT

OVERRUN

OVERRUN

COMPLETION

NUMBER

8- ORIG COST

2(3x7)

SUM 8

3/1 "2

5x10

6+11

PCAC

! ': ' ! i

• i ;

0.00

0.00

0.00

0.00

0.00

20.00

REPORT 1

500

10.00

10.00

2.50

20.00

45.00

REPORT 2

5 00

10.00

20.00

■> •■•■>

15.56

40.56

REPORT 3

c r.r,
J.U'J

4 A ft ft

iu.UU

30.00

1 .00

4 4 -.c

36. ij

REPORT 4

5.00

10.00

40.00

1.80

3.00

33.00

REPORT 5

z r.r.
J.U'J

4 r, .->.-.
: IU.UU

en nn
JU.UU

4 r,r,

1 .00

z zr
J.JO

•-\c zr
iU.UO

r. r- !-.■—. r.-r r-

ncrunl o

5.00

10.00

60.00

1.22

j'.6r

^3.67

REPORT 7

j.uli

« .". r.r,

1 u.Uu

->.-■ n,-.
( U.'JU

1.03

i\ 4 r,
<.. 1 O

r .-5 4 r,

ir .id

ntruni o

5.00

10.00

80.00

0.33

33

25.33

REPORT 3

u r.r,

a.uu

4 a ft n

1 U.UU

OU.UU

.-> r.r,
U.JU

■" r.r,

U.UU

AC AT.

-J. 'JO

i-.r-T-,,— r-,-r 4 ■-.

ncrurj i ■■-

Figure ".14 - Im| ivement Analysis

~\

COST TREND WORKSHEET

■y
■'.tV.'V.

CDST SCOPE > FACTOR SET AT M'-. ANG Cut \$ FACTOR xAM : : -x

ESTIMATED

,>S*S : : : : : :

: : : : : : : : : x : tOx : : : x : x:
COST

PREDICTED

PREDICTED

rn-"T at

acpnRT

x-x-j

OVERRUN

c_ HQir; prior

2*(3x7) SUM 3 '9/(1*3)

- f 1 a L — („

iVbHrtUN

Rv 1 n

LUMI-'Lt I iUN

.11

NUMbtH

~l

ALI A - PCA.C

U.00

u.uu

u.uu

R ftfl

1 n nn

1 i*i on

OJ/O
1 9K

U.UU

iU.UU

1 n nn

vp. nn

REPORT i
ocpnoT'T

5.00
5.00

1Q.00
10.00

20.00

0.74

5.13

30 UU

0*7

.81

30.13
27.31

REPORT 3
REPORT4

5.00

Too"

10.00

1 n nn

4U_UU_

50.00

0.32
23

1 60

U.3;

26.60
25 33"

R EPO RT 5

o.uu

1U.UU

bU.UU

0.17

U.Dii

REPORT

5 00

1 90 ?Q OQ

n 14

n 27

25.27

REPORT 3

1

;

5 00

1 0.00 80.00

O.J i

0.11

25.1 i

REPORT 3

1
1
1

5 00

10 00 30 00

0.09

fi : J M

25.00

REPORT 1

1

COST SLOPE FACTOR

SET AT Z#

W*0COt ■■» FaCK

tf At 4. .... . .

|

•••I

ES i iMA i ED

COST

PREDICTED

PREDICTED

COST

SLOPE

COST AT

REPORT

OvERRun

OVERRUN

COMPLE 1 ION

NUMBER

1
1

fi- QB1Q COST

3"! 3x71 SUM 3

; — t ■ w /

5x 1

6+11

ALT B - PCAC

!

i

1

1

0.00

0.00 0.00

0.00

0.00

20.00

REPORT 1

1

5 '"'0

1 5 00 1 5 00

1.88

1 p. nn

40 00

REPORT 2

1

5.00

15.00 30.00

1.11

: 7 .78

32.78

REPORT 3

!

!

5 00

15.00 45 00

0.70

4.22

?q W

REPORT 4

5.00

15.00 60.00

0.43

2.40

27.4(1

REPORT 5

1

5.00

15.00 75.00

0.35

1 .33

26.33

REPORT 8

1

j

5.00

15.00 30.00

0.2S

0.73

25.73

REPORT 7

1
1

5.00

1 5 00 1 05 00

9.21

0.41

25.41

REPORT 3

1

1

5.00

15.00 120.00

O IS

0.16

25.18

REPORT 3

i

^ 00

15.00 135.00

OH

00

25.00

REPORT 10

I

1

Figure Z.\\ - Improvement Analysis

P:\j-- S3

3.6 MACRO PROGRAM FOR EXCEL

A macro program was developed on an IBM Excel Spreadsheet.
This macro was developed in conjunction with this research to aid
the novice user of this forecasting method in the use of the
spreadsheet and chart. The macro utilizes the customization
assets of the Excel Spreadsheet to form interactive programmable
boxes to guide the user throughout the initiation, periodic
updating and printing of the results. The macro program is
detailed in Appendix D.

66

CHAPTER IV RESULTS

In Chapter III a detailed analysis of Davis's spreadsheet
was presented. Several different scenarios were discussed as
well as a detailed sensitivity analysis. The different results
as discussed in the last chapter are summarized below:

1. The spreadsheet is easy to learn, use and modify.

2. The different scenarios used to analyze the spreadsheet
all showed the special convergence feature of this process. The
convergence of the ECAC to the PCAC reflects the idea that
changes in the ECAC in the beginning of the project are more
critical than the same changes at the end of the project.

3. The different scenarios also showed the amplification
factor of this process, this amplification factor is defined as
the difference between the ECAC and the PCAC. This factor is
extremely noticeable in the beginning of the project and is
reduced as the project progresses. Amplification factors of 180%
are common in the beginning whereas they reduce to at the
completion of the project.

4 . The macro developed in the research is extremely easy to
use and can be given to a novice computer user to initiate the

5. The basic theory behind this spreadsheet is that changes
in the ECAC will be continuous throughout the remainder of the

67

contract. These changes therefore should be calculated and then
multiplied by the remaining portion of the contract to determine
the PCAC.

6. The spreadsheet does not function when faced with an ECAC
that shows large decreases, the PCAC due to the amplification
facto will reflect negative values.

complexity and will require additional programming of the
equations. This increase in complexity will be difficult to
learn for the novice computer user and will decrease the user-

68

CHAPTER V RECOMMENDATIONS

From the results section of this report several
recommendations can be provided in the use of this worksheet.
First, the spreadsheet should be customized to the individual
company's requirements. The amplification factor should be
analyzed and the factor should be discussed and agreed upon prior
to using the spreadsheet. Along with the amplification factor
other customizing factors should be considered including the cost
slope and whether to include several different PCAC's on the
chart .

This spreadsheet should not be used alone in determining
courses of action due to cost overruns. This process is only
used to identify that a problem exists, it does not identify the
location of the problem. Other methods discussed in Chapter 2
should be used in addition to this process to determine the
location and the extent of the problems.

Future research in this area should include the combination
of the methods described in Chapter 2 and this method and then
programming this combination with the cost accounting of the
project to derive a final predicted completion cost which can
identify problem location and extent. The references provided in
the Reference section of this paper will aid any new research.

69

CHAPTER VI CONCLUSIONS

The main objective of this research paper was to analyze
Davis's spreadsheet and chart to determine if this method of
forecasting was reasonable and practical in the construction
industry. Several currently used forecasting methods were
discussed so that a comparison of Davis's method could be made.
The equation and spreadsheet were discussed in detail and then a
sensitivity analysis was performed on the spreadsheet and chart
to determine the cause and affect of different scenarios.

After the analysis of the simulation, several factors were
identified in the spreadsheet that required changes. One of
these changes was the amplification factor which is identified as
the difference between the ECAC and the PCAC. Another change
suggested was that the converfence of th ePCAC to the ECAC by
accelerated. Both of these factors were combined to improve the
practical.

In the results section of this paper several items were
discussed including the ease of use of this particular
forecasting method. Another advantage of this method is that it
is easy to manipulate and therefore a project manager can easily
change the equations within the spreadsheet so that the results
can better show the trend of the project to date.

70

Another major problem of this spreadsheet is that the
amplification factor is to large. This factor, if used too often
can be a detriment to the project manager. If every little
change in the ECAC produced a large change in the PCAC, the
manager is more likely to look upon the PCAC as just another
annoying factor of his job and will eventually ignore the warning
signs. One way to reduce this problem is to limit the use of the
forecast method to the last 80% of the project life, the
amplification factor gets smaller as the project progresses.

The convergence of the PCAC into the ECAC curve is an
interesting aspect of this forecasting method. This convergence
aspect shows that differentials early in the project's progress
are more important that the same differentials in the later
phases of the project. This part of Davis's equation is
realistic, an estimate that has large cost differences in the
beginning of the project is more likely to have additional cost
differences throughout the term of the project thus compounding
the problem.

The underlying theory of this forecasting method is that a
predicted cost at completion can be determined by analyzing
differentials in the estimated cost at completion during specific
times in the process of the project. This theory would be
extremely hard to sell to any construction manager. Variance
analysis of cost codes provide the same information without
enlarging the variance due to the time of the estimate. The
spreadsheet as presented is unreasonable and should never by used

71

as a sole source of forecasting.

The formula used in conjunction with other methods however
could be a valuable tool in alerting managers to future problems
Other methods such as variance analysis, productivity curves and
the managers experience could all be used together to establish
an effective forecasting method.

72

REFERENCES

Davis, J. G. (1975). "Keeping Project Costs in Line." Machine
Design . Dec 11, 1975, 128-133.

Clark, F.D. and Lorenzone, A.B. (1985) Applied Cost Engineering .
2nd Edition, Marcel Dekker, Inc NY, NY.

1987 Transactions. American Association of Cost Engineers, 31st
Annual Meeting . Atlanta, Georgia

Stevens, W. M. (1984) . "Cost Control: Integrated Cost/Schedule
Performance." Journal of Management in Engineering . ASCE, 2(3),
157-164.

Riggs, L.S. (1987). "Project Control Techniques.", Project
Controls; Needs and Solutions . Edited by C.W. Ibbs and D. B.
Ashley, Prodeedings of a Special Conference Sponsored by the
Construction Division of ASCE, Chicago, 111, June 8-9 1987.

Bessa, J.L. (1983) . "Cost and Schedule Controls During
Construction . " , Current Practice in Cost Estimating and Cost
Control, Edited by V.F. Sporull and C. Popescu, Sponsored by the
Construction Division of ASCE, Austin, Texas, April 13-15 1983,
ASCE, NY, NY.

McMackin, P.J. (1985). "Control of the Design Process.", Making
Project Control Systems Work . Edited by P.M. Teicholz, ASCE, NY,
NY.

73

APPENDICES

A. COST TREND CHART WORKSHEET OPERATION MANUAL

B. DAVIS'S ARTICLE

D. EXCEL SPREADSHEET AND CHART MACRO

74

APPENDIX A

COST TREND CHART WORKSHEET OPERATION MANUAL

COST TREND CHART WORKSHEET
OPERATION MANUAL

INTRODUCTION

In 1976 Gordon Davis developed a Cost Trend Chart and
Worksheet that would calculate the Predicted Cost at
Completion (PCAC) when provided with periodic updates of the
Estimated Cost at Completion (ECAC) . This worksheet
computes the difference between the Estimated Completion
Cost (ECAC) inputed for the current period and the input for
the previous period. This difference is then multiplied by
the number of remaining periods in the project to determine
the Predicted Completion Cost (PCAC) . Caution must
excercised in the use of this spreadsheet and chart, other
methods of trend analysis and forecasting must be utilized
to formulate the final completion costs. This equation
tends to yield early results that some may consider rather
large completion costs due to minor differences in the
estimated cost at completion.

The spreadsheet and chart are provided in the next few
pages. The spreadsheet is rather easy to intiate, first the
title and original estimated cost are inputed at the
beginning. Periodic updates only require that the update
number, projected completion date, and the estiamted cost at
completion be inputed. The final result is then
automatically calculated through the equations and listed in
Column 12, the predicted cost at completion.

The chart is simply the graphical representation of the ECAC
compared to the PCAC. The horizontal axis is the Cost
Report Date and the vertical axis is the completion cost.
The input and the result is plotted for the purpose of
comparing the ECAC to the PCAC.

The following is a column by column breakdown and

A. INITIATION OF THE WORKSHEET

1. TITLE: this space is provided to input the title of the

construction project and any pertinent comments.

2. ORIGINAL COST: the original estimated cost at

completion is imputed into this
worksheet block.

B. DESCRIPTION OF THE COLUMNS

REPORT NUMBER

Prior to column one, the worksheet presents a column showing
the report number. In the analysis of this worksheet these
updates are stated in monthly intervals. However the
updates may be in any interval as long as the unit of the
interval is consistent.

COLUMN 1: REPORT DATE

This column is a restatement of the previous column. The
value of this column is used later in the spreadsheet, the
word "REPORT" is therefore eliminated in order for the value
to be used in the spreadsheet program. Similar to the first
column this column shows the report date as a monthly
update .

COLUMN 2: PREVIOUS REPORT DATE (INPUT)

This column simply states the last report date and is
inputed as the number of the last report date.

COLUMN 3: UPDATE INTERVAL

The update interval is the difference between the first
column and the second column. This shows the amount of time
that the updating of this spreadsheet has lapsed.

COLUMN 4: PROJECTED COMPLETION DATE (INPUT)

This value is imputed into the spreadsheet as the estimated
time to compete the project. Note that this value is part
of the periodic updating of the spreadsheet. This value
represents the total duration of the project, not the
remaining completion time.

COLUMN 5: REMAINING TIME TO COMPLETE

This column is automatically calculated by subtracting
COLUMN 1: REPORT DATE by COLUMN 5: PROJECTED COMPLETION
DATE.

COLUMN 6: ESTIMATED COST AT COMPLETION (INPUT) (ECAC)

This input value is the most up-to-date completion cost
estimate of the project. Usually a project manager uses a
combination of the methods described in the previous chapter
to determine this value.

COLUMN 7: ESTIMATED COST OVERRUN

Column 7 represents the estimated cost overrun for the given
interval. This calculation is simply the difference between
COLUMN 6: ESTIMATED COST AT COMPLETION and the ORIGINAL COST
as imputed during the initiation phase of the setup.

COLUMN 8: CALCULATIONS

This calculation consists of multiplying the product of
COLUMN 3: UPDATE INTERVAL and COLUMN 7: ESTIMATED COST
OVERRUN by a factor of 2.

The reasons why the factor of 2 is applied is not apparent,
in the sensitivity analysis section of this chapter this
factor is discussed.

COLUMN 9: CALCULATIONS

This column is the summation of the previous column to the
point of the update. For example, if the report date is
number 5, then the value of this column would be the
summation of the first five entries of COLUMN 8 .

COLUMN 10: COST SLOPE

The cost slope is derived by dividing COLUMN 9 by the square
of COLUMN 1: REPORT DATE. As can be seen from the graphical
representation of this spreadsheet, the cost slope is the
slope of this line between the two update intervals.
Without the square root of the denominator this equation is
basically the average of the interval cost overruns.

In the sensitivity analysis of this spreadsheet formulas the
discussion of why the denominator is squared is discussed.

This value is calculated by multiplying COLUMN 5: REMAINING
TIME TO COMPLETE by COLUMN 10: COST SLOPE.

COLUMN 12: PREDICTED COST AT COMPLETION (PCAC)

The final column represents the results of this spreadsheet
and is calculated by the addition of COLUMN 6: ESTIMATED
COST AT COMPLETION and COLUMN 11: PREDICTED ADDITIONAL
OVERRUN.

ZCST TREND WORKSHEET

.•• a- ; :- : -::-z:::-"-7v" •"■".:•.' :, ::.i: ,.:::., ii'i: : ."i i. .:....: n,../ ... ■. *

cost mem worksheet :.' j

mvisornGMJiLAmmz

DATE 25-May-9Q

■

ORIGINAL COST= j 5

1 L *l J O

PREVIOUS; PROJECTED REMAINING ESTIMATED

REPORT i REPORT i REPORT ; UPDATE COMPLETION TIME COST AT

NUMBER DATE DATE : INTERVAL ! DATE j TO COMPLETE ; COMPLETION

INPUT CCL4-CCL1 INPUT

ECAC

."JGINAL 6 5.00

: PORT1 10 16 5 5.00

: PORT2 2 118 8 5.50

F PORT 3 3 2 1-3 5 5.50

F PORT 4 4 3 1 13.09 9 09 8.00

FPORT5 8 4 2 3.77 3.77 8.50

FPORT6 7 8 1 3.87 2.67 3.00

FPORT7 8 7 1 3.75 1.75 3.00

PORTS 8 8 -8

PORTS 9 9 -9

PORT 10 10 3 1 -10

7 i 9 iu 11 12

ESTIMATED COST PREDICTED PREDICTED

COST SLOPE ADDITIONAL j COST AT REPORT

OVERRUN OVERRUN I COMPLETION i NUMBER

b-ORiGCOST 2(3x7) SUMS 9/1*2 5x12 6+11

PCAC

0.00 0.00 0.00 0.00 0.00 5.00 REPORT 1

050 1.00 1.00 0.25 150 7.00 REPORT 2

0.50 1.00 2 00 0.22 1.11 8.81 REPORT 3

100 2.00 4 00 ! 0.25 2.27 8.27 : REPORT 4

350 14.00 18.00 0.50 1.89 10.33 REPORTS

400 8 00 26.00 53 1.42 10.42 REPORT 6

4 00 8 00 34 00 0.53 0.93 9 93 REPORT 7

-500 0.00 34.00 0.53 -4.25 REPORTS

-500 0.00 34.00 0.42 -3.73 REPORT 9

-500 -10.00 24 00 0.24 -2.40 REPORT 10

INDEPENDENT RESEARCH PROJECT

DAVIS ORIGINAL ARTICLE

12.00 -r

10.00 --

8.00 --

COSTS 6.00 --

K.

/

/

;

4.00 --

2.00 4-

0.00

/

*

*/

H 1 1 1 1 1 1 1 1 1

123467889 10
MONTH

APPENDIX B

DAVIS'S ARTICLE

J. GORDON DAVIS

Senior Partner

Systems /Project Management

Atlanta, Georgia

Keeping Project
Costs In Line

For most projects, profitability depends
upon effective cost control. Unfortunately,
plan and analysis discussed here should
go a long way toward eliminating
these difficulties.

Many engineering managers try to keep on top
of project costs by merely reviewing cost reports
as they become available. This procedure is forever
lagging behind what is actually occurring. What is
needed is a method for forecasting costs and
completion dates so that efforts to keep costs in
line can be launched before things get out of hand.

It's The Trend That Counts

The typical cost report contains a breakdown of
project estimates into categories, as shown in Fig.
1. The cost figures presented for each line item
time the project was funded. The Current Estimate
frequently includes only the Original Estimate plus
any cbtnge ordfy K '- -"?r rr % -'' ' ' ■■ to label t u; :

Currc-.i- Estima;. ;uov*..".g ^ = moa. .'eat estimtuc,
funded nr nnt nf what thP itom should cost. Actual

The Actual-To-Date figure should include all
charges for work accomplished to date, not just
figure includes all contractural obligations for fu-
ture delivery of work or products. The Estimate-To-
Complete figure is the current estimate for the
work neither completed nor under contract

The Expected-Total-Cost figure is the sum of the
Actual, Committed, and Esdmate-To-Complete fig-
ures. Its deviation from the Current Estimate is a
measure of performance." Its deviation from Funded
Estimate is a measure of profitability, although
many managers try to use this deviation figure to
measure performance. Its deviation from Original
Estimate is essentially meaningless. Yet, it is this
comparison which some managers use exclusively.

Many cost reporting systems omit the Estimate*
To-Complete figure and, thus, give little indication
of an impending cost overrun for a line item. Such
systems may nevertheless be of some use if a large
number of line items are involved because only »
small percentage of the line items are ever in prog-
ress at one time.

Looking at one or even a series of cost report>
will not give a good picture of what costs can be
expected for a project, even if the report totals u?
the expected total project cost It is the trend in
this total cost figure which gives a basis for fore-,'
casting the project cost at completion. !

' r '-- '"rpect-e-i -~ •■*■ »l-Cost grar^ >" a ti

■><-..!

dfc .... fuild proj ,;> shown i. . .£. £■•<■ ■^'•'•'
first few project cost reports came throuph tr

H30ST STATUS REPORT

XYZ CORPORATION

ALPHA PROJECT

)

(|>RY ORIGINAL CHG ORDERS

CURRENT ACTUAL COMMITTED

ESTIMATE EXPECTED
TO COMPLETE TOTAL COST

DEVIATION

FROM
CURRENT
ESTIMATE

tal
e

13.500

850

14 350

5 150

2 500

8000

15 650

1.300 +

pical cost report is a record of various esti-
tning costs for each line item of a project,
lortant indication of a project's performance

provided by such a report is the deviation of Expected Total
Cost from the Current Estimate.

Actuol total cost ot completion -

I icted totol cost as ot 6-2-74

Actual completion date-

Predicted completion date os of 6-1-74
(from completion date trend chart)

-L.

s

^UU

10 5-5 6-30 8-25 10-20, 12-15; 2-9 ( 4-7
I 4-7 6-2 7-28 9-22 11-17 1-12-75 3-7
Cost Report Date

npomnce of a forecasting procedure is illus-

Expected-Total-Cost graph of an actual design/

:(a). - Instead of merely plotting cost figures

I reported, the manager of this project could

J the cost trend (dotted line) inherent in those

11-1-73,1-1-74 3-1 5-1 j 7-1 9-1 11-1 ,1-1-75: 3-1 | 5-1
12-1 2-1 4-1 6-1 8-1 10-1 12-1-74 2-1 4-1
Report Date
figures. By applying the same projection procedure to the
scheduled completion date of the project (b), he would have
had an early warning of the cost problems that were build-
ing. Instead, his reports were always behind what was
actually happening. j i

i was strictly an engineering project at
id the word soon came back from engi-
: costs were under control. But the graph

ire, it might have predicted a trend line

dashed line shown on the graph. How

the trend line be projected? To the

ipletion date, of course. But the project

date is having troubles too, Fig. 2b. If

continues, the project completion

:cur until the trend line crossed the iso-

is this project completion date to which
end line should be projected (see Fig.

lalysis is crucial to controlling project j
nds to remove the bias of tight or loose
low or high cost estimates, and good or
m8BB4 Th e typical result of trend analy-
st management disturbed much earlier
herwise would be. This early warning
lanagement to invoke control measures
;at number of options are open.
step procedures for creating proper
t-date and total-cost trend charts are de-
the box entitled "Predicting Principal
rameters."

Down Total Costs

ring-cost overruns are so commonplace
is usually no uproar until the percentage

1. 1975

overrun reaches at least double-digit proportions.
An attitude of "we'll just have to live with it" is
frequently a cover for uncertainty as to whether
there is something wrong with engineering per-
formance or whether the estimate is bad, or both.
This attitude may also indicate uncertainty as to
what actions can be taken if the problem stems
from poor engineering productivity. Engineering
cost control also suffers from the fact thaf engi-
neering costs are typically only a small percentage
of total project cost.

The magnitude of engineering cost overruns is
all too often underestimated because the ripple ef-
fects are not included. If an engineer takes 50 hours
more for a task than estimated, the increase in cost
will be insignificant on a million-dollar project.
However, an increased number of man-hours is
frequently the result of a change in a preliminary
design feature. Such a change can lead to changes
in the cost of materials and equipment, the cost
of design modifications on related portions of the
system, the cost of producing or constructing the
system, and/or indirect costs which vary with
project length.

While these changes could be cost reductions,
they usually are cost increases. The engineer is
much more likely to beef up the preliminary de-
sign than to trim it down.

neering* changes can, bo accomplished through thek

use of work breakdown packages^ These packages
can be used from the proposal stage through project

129

Predicting

Principal Project Parameters

Completion Dates

The Comp.etion Date Trend Xh.it is
set up with one calendar date
scale on the bottom ax.s and
Kher on the .eft-han. ax,, £ a .
dailv or weekly gnd. ine bw
P„da, should start «*.£
Hate of the proiect. This scaie
should be labe.ed Update Date

initially scheduled for proiect

1 Date An iso-date line is then
Srawn. This is the straight line
"passing through all points MFor wh.cn
ooth calendar scales gwe the same

date. This 45-deg line S the
tareet line in the sense that trena
SSTJai be projected to the.r
intersection with the iso-date line.
The points to be plotted on ^.s

ulrt are the Earliest Scheduled

activities being tracked. Each
entry is made by finding, on the

bottom axis, the date on wh.ch the
,atest schedule update was
accomplished, then movmg vert.cally

to a point which shows the
completion date resulting from the

♦« hP revised — that is, actual
Sog e« exttTy matched the initial
schedule-the completion date
trend line would develop as a
horizontal line. This line , wou £
intersect the iso-date «'"••" *^,
day that the milestone act.v.ty was
completed. If progress towarf
that milestone was slower than

scheduled, the line would **™*
positive slope. Conversely better
than-scheduled performance would

%"/ 5

result in a negative slope to the
trend line. Regardless of the 'evel
of performance, the trend I line
can be projected beyond the last
pto^P^int to an intersection w.th

the iso-date line. The po.nt of
Intersection has a value on the
Completion Date axis which i may
be interpreted as the Pred.cted

'Tsutr^fthat the performance
.^whether high or low re.at.ve
to scheduled performance, ^ .s a
constant. This means that the

trend line is straight and that a
projection can be made by s-mply

Sending the trend line. Thus.
a project on which t.me est.mates
were too low will show a trend
,ine with a positive slope, and the
projection will result in a predicted

Completion date greater than the

'/ 7 V 15 \z V 29 % Al

Report Dote

currently scheduled ear .est
completion date. S.m.larl* a
project on which time eshmates
were too high will develop a trend
line with a negative slope and t*
predicted completion date wri I be
earlier than that result.ng from

the most recent update.

The visual projection of the trend

line formed by the Earliest-
Scheduled-^mpletion-Date pom*
is subject to a great deal of
variation, To standards th.s
projection, the following procedure
should be used:

1. Assign week numbers start, g
with zero to each end-of-week
point on each axis. Let X equal
the week number on he Report
Date axis and Y equal the week
number on the Complet.cn Date
axis.

than-scheduiea pww»— —

COMMON DATE TREND CHART WORKSHEET

•. " 5

Yong

i

2.

Trent
3.
new
Colui

4.

I

)ject Cost Trend Chart

e:ted cost at completion

Project completion
date predicted
as of 12/24

stimated cost at completion

I

! I

2 3 4 5 6 7 8 9 10 11 12 13
"/,2 % "/a; ' 2 /3 12 /10 12 /17 12 /24 ,2 /31 W V, 5 V 22 V 29
Cost Report Date

i Completion Date
rksheet as shown,
update, start a
row by filling in
4.

he entries for each
rtn in numerical

12 is the projected
t expressed in work

start of the job.
;r in Column 12 may
le Completion Date
assist in its

lost Trend Chart
i calendar data
ottom axis and a
le left axis of a

rectangular grid. The calendar
date of the project. This scale
should be labeled Cost Report Date.
The cost scale should start
approximately 10% below the
initial estimated project cost. This
scale should be labeled Total
Project Cost.

The points to be plotted on this
chart are the sum of Actual
Costs to Date plus Committed
Costs To Date plus Estimated
to these points as Estimated Costs
at Completion. These data are
taken directly from the Job Cost
Status Report (See Fig. 1).

be joined by straight lines to
make it easier to visualize the
trend of these points. If the

original estimate and the project
execution were both perfect, this
line would be horizontal. However,
many factors tend to cause the
Estimated Cost at Completion to
rise from one cost report to the
next. Projection of this trend to
the anticipated project completion
date will give a Predicted Cost at
Completion. The anticipated project
completion date should be the
Predicted Completion Date from
the Completion Date Trend Chart,
rather than the Earliest Scheduled
Completion-date — which could occur
only if all factors causing project
slippage were suddenly eliminated.

The visual projection of the cost
trend line is subject to a great
deal of variation. This projection
use of the accompanying Cost
Trend Chart Worksheet. The
following procedure should be used:

1. On the Cost Trend Chart,
assign week numbers starting with
zero to each endofweek point on
the Report Date axis. Let X equal
the week number on this axis.

Let C equal the cost figure on the
Total Project Cost axis.

2. After each update on the Job
Cost Status Report, start a new
worksheet row by filling in Columns
1, 4, and 6. Column 12 of the
Completion Date Trend Chart
Worksheet is the source of Column
4 of the Cost Trend Chart
Worksheet. If a cost figure

is generated for a date on which
no schedule update has been made,
the most recent Predicted
Completion Date should be the
Column 4 entry on the Cost^Trend
Chart Worksheet. lS '

3. Calculate the entries for each
remaining column in numerical
order. Column 12 is the predicted
project cost at completion. This
figure may be plotted on the Cost
Trend Chart to assist in its
interpretation.

1ART WORKSHEET

2

3

4

5

6

7

8

9

10

it

12

PREVIOUS

REPORT

OATE

UPDATE
INTERVAL

PROJECTED

COMPLETION

DATE

REMAINING

TIME TO
COMPLETE

ESTIMATED

COST AT

COMPLETION

ESTIMATED

COST
OVERRUN

2|3 « 7|

S 8

COST
SLOPE

PREDICTED

OVERRUN

PREDICTED

COST AT

COMPLETION

1 -2

(INPUT)

4 - t

(INPUT)

6 - C ong

9 - 1»

5 x 1C

6 - ii

60

50

1

60

50

50

00

00

00

00

00

50

1

1

80

60

55

05

1

I

25

i 5

•o

2

1

(8 0)

50

55

05

I

20

22

i i

66

3

t

1309

909

60

1

:o

40

25

2 27

8 2*

4

2

9 77

3 77

85

35

14

180

50

i 89

10 39

6

1

9 67

2 67

90

40

80

26

0?3

1 '2

'0 J2

7

1

9 75

1 75

90

40

80

34

53

93

9 93

4>

J75

131

low Well Do You Manage?

>u can diagnose how well you manage your
ojects by the pattern of frustration you
:perience over the life of a given project. The
10 basic patterns of frustration can be illustrated
t the extremes of loosely-managed and tightly-
anaged projects. The first curve in the graph
lows the pattern of a loosely-managed project,
uring the early stages — usually the period in
hich most of the engineering takes place — the
lanager lives in a condition of uninformed
ptimism. A false sense of well-being exists
ecause there is no data indicating anything amiss,
his period draws to a close with the appearance
If a vague concern. Here the project manager
tarts issuing serially numbered memos in a
rocess known as "cover your rear".

As engineering winds down and production
r construction gets underway, the frustration
limbs at an increasing rate. The cost and schedule
lata are beginning to filter through and the result
> the start of an effort to "search for the guilty",
ihis marks the start of the terminal phase,
ailed "panic". As deadlines are missed and
:ost overruns become real, the manager usually
ndulges in the final rite of "punishing the
nnocent". At this point traumatic changes in
irganization take place, and the project is made

None

Proiect
Stort

to wallow to a conclusion — late, overspent, and
disruptive to the organization.

On the other hand, if a project is tightly
managed, the manager will be keeping track of
schedule and cost trends. If the initial data is
bad, the manager's early informed pessimism will
lead him to take action while many options are
still open. If these actions are successful, his
frustration will peak as the flow of new problems
drops below the solution rate on the problems

etion to give a uniform basis for providing
urrent Estimate.

starting with an estimate which goes to a
table level of detail, a basis is established
:king up cost changes. Each engineer involved
i design can be assigned tasks which relate
specific set of one or more work packages,
packages will carry cost estimates for di-
mgineering costs, materials, and equipment
ase prices as well as direct construction costs,
package will be estimated in terms of unit
ities and unit prices. Thus, an engineering
:arries with it the requirement for quantify -
I design decisions in terms of the cost deter-
ions which result.

se cost determinations must not only include
;ering, procurement, and fabrication, but must
over those same factors on work packages
are affected by the package in question. The
ier releases a work package only after sum-
ng the total cost change which has resulted
lis work on that package,
advantage of this approach is that the de-
has had a goal to shoot for, which is im-
it in at least two waysJfRrstJ he is forced to
ne the preliminary or proposal stage design
effort to stay within its cost boundaries; thus,
edback to the estimating system is much
precise than would otherwise be the case,
requently, the engineer strikes out afresh to
t his portion of the project, taking the atti-

tat the ori? ; -'

do:

merely to allow an estimate to be made. Second,
the knowledge that a certain number of man-hours
have been budgeted for his activity helps the engi-
neer determine the quality level that can be af-
forded on this activity.

The engineer's release of a work package will
be followed by its tentative updating in the cost
control system. A threshold can be established such
that increases of more than, say, 5% over the
previous Current Estimate are singled out for man-
agement review. Alternatively, the review threshold
could be an absolute amount of cost increase in
all related work packages. Management then has
the option of rejecting costly design changes be-
fore they are irretrievably assimilated into the to-
tal system design.

The completion of work packages within the
estimated man-hours is dependent upon the ade-
quacy of the estimate. Estimating standards which
are too tight create an atmosphere in which the
futility of staying within the estimate leads to an
attitude of ignoring the estimate, because it is ob-
viously invalid. If the tightness of standards is the
result of a managerial decision to set high goals.
an improper and ineffective use of the cost con-
trol system results.

Standards should represent expected values, sue.
that the sum of all the estimates for each work
package will lead to the expected project cost. !•
management decides to compete with other organs
zations for a project, and wishes to bid a '" v
"it^! nri' norms' Tit'.rrate s'*-"

exactly as it w
profit margin,
petitiveness of
fined to the siz
a conscious dec
of work perforr
Estimating st
loose because f
ance is not avai
ing of the stan
independent of
cannot. The e.'
what the organi
expect for the i
formance throug
improvement, r
should not be i
tern unless actu

Scheduling F<

Assuming thai
pineering cost c<
performance wh
ards. Whether w
life or the life o;
package, it is im
of the work earl}
the work for the
work in interval:
man-hours by 01
to consist of ph;
end points. For
broken down as f
• Planning — 0.
Review prop
Preliminary
Calculations-
Review of si
Final sketchf
Specification
With this kind
better track his
completion goal,
exert control by
each day, giving ;
pated during th
achieved during
v 'sor never rejec
'ual performance
went will norma

tivity.

One factor that
u| ing the 20 man
^an-hours av?ilaV

tli as it would for a project bid at a normal
loargin. The managerial decision on com-
j-ness of the bid should have its effect con-
I ) the size of the mark-up for profit, unless
r:ious decision is made to lower the quality
k performed below existing standards,
liiating standards may be too tight or too
»>ecause feedback reflecting actual perform-
. not available in a form that leads to updat-
( the standards. Unit costs can be updated
pdent of performance, but unit quantities
1 The estimating standards should reflect
the organization's performance leads one to
I for the next project. Improvements in per-
:ice through procedure analysis, management
r'/ement, motivation, or any other factor
i not be anticipated by the estimating sys-
nless actual performance data is available.

duling For The Short Haul

uming that good standards are in effect, en-
ing cost control then focuses on obtaining a
•mance which meets or betters these stand-
Whether we are talking about overall project
r the life of an engineering task on one work
ge, it is important to identify the components
; work early and determine the implications of
ork for the immediate future. In engineering

this means the short-range scheduling of
in intervals of as little as one-half hour. An
eering task having an estimate of, say, 20
lours by one man should be planned so as
nsist of phases or activities with identifiable
joints. For instance, such a task might be
n down as follows:
Manning — 0.5 hr
leview proposal — 1.0 hr
Veliminary sketches — 6.0 hr
-alculations — 2.0 hr
Review of supply catalogs — 1.5 hr
'inal sketches — 7.0 hr
Specifications — 2.0 hr

th this kind of breakdown, the engineer can
' track his own progress toward the 20-hour
letion goal. The engineering supervisor can
control by having the engineer report once
toy, giving a statement of the progress antici-

during the next day and the progress

"ed during the past day. Even if the super-

never rejects the goals set or criticizes ac-

Pwformance, making a short-range commit-

•ffl normally improve engineering produc-

s factor that must be accounted for in sched-
ule 20 man-hours of work is the number of
tows available per day for unscheduled work.
*ring the telephone, responding to questions

from associates, handling correspondence, and other
similar tasks frequently consume much of the en-
gineer's day. In practice, it is advisable to assume
that some negotiated figure, say 60%, is the proper
factor to represent available time. Thus, one would
expect the engineer to average 4.8 hr per day on
scheduled production. This figure can be revised
after the short-range scheduling system has been
in effect long enough to be debugged.

Work sampling by the engineers themselves can
lead to a refinement in the estimate of nonsched-
uled work load. For instance, each man can tally
his activities into scheduled or nonscheduled cate-
gories every half hour for a week. The category
percentages will be good estimates of the total
week's time distribution. Again, the effect is to
make people more aware of the degree to which
nonscheduled activities are allowed to interrupt
scheduled work. The engineer may decide to group
his nonscheduled activities to some extent, reduc-
ing the effects of interruptions and shifts in type
of activity.

Assign A Project Manager

A project is frequently allowed to wander
through the various functional groups involved
without a real advocate who is in close touch with
current status and has the authority to make ex-
pediting decisions. The project manager should be
assigned during the proposal stage, even though
the majority of such assignments will not lead
to funded projects. This timing avoids the prob-
lem of having an accepted proposal handed to a
project manager who immediately begins to find
flaws in the concept, the estimate, and^the sched-
ule.

The project manager must be given the oppor-
tunity to set up the control system which allows
him to accurately determine project status. He
must be supported by an estimating system which
is aligned to the realities of the organization and
keeps the estimate revised to reflect current knowl-
edge. He must also have the support of top man-
agement so that he is a true project manager, not
merely a coordinator w v ith no authority.

Engineering cost control is not likely to succeed
unless the system is thoroughly planned and de-
bugged before implementation. Computerized es-
timating systems are necessary if responsiveness
to design changes is to be adequate and feedback
for adjustment of standards is to be thorough. The
systems should be designed so as to require as
little input effort as possible from the individuals
carrying out the activities. The rewards for pay-
ing the initial price are large, with payout on initial
costs, frequently being as short as six months into
the first project. □

ihtr II

APPENDIX C

o

<

O

LU

<

DL

n i
r t
I I-

/ i

LU

4- od

x

LU

I-
CO
LU

LU

to

<

LU

LL

o

p

p

rf

n

/

rt

t
o

4- u-

T

4- m

t w

■_u

■TJ"

□

a
10

LO

a
a

m

a

a

| —

C

■—■

a

CM

□

CM

to

<—•

in

1*?

3 J

[>:•:•

tvXvfcv

t>>:x*5>:

II

tXvXgx
!:•*•'
£x\$£

H

x : xftx
r V

m

XS«Q:

m

m-

m
:*a
■m

i

.V.V.V.V.V.V.VJrS

1 I

:0|
CJ

:_l:

!<!

Z

in

:co

id

n

uu

CO

D

h-

, ,

UL

UJ

1—

•->

<

UJ

III

L-

'X

IT

u-

CJ

w

cm

31

CM

CJ

CM

n

-

CJ

m

ul

CM

to

IN

173

CJ

CM

0)

CM

GO

N

CM

CM

03

!-■-

O : O

in

UJ

!

O 1

<

< 1

LU

Q.

S

n

u

1

T

I

GD

EL
LU

Q

LL

LU

G

<
I
n

~*.

O

Cl

f-

LU

h

00

LL

U

LU

H

— 1

2

*^

T"

I

'J

DC

n

U

—

LLl

n

>

^u

L)

:-

LL

<i

Q.

n

u_

u.

r

>.

— '

Q

z

<

iz

I u

>

4- lo *"

Ck

^

in

Cl

LO

LO

h

rXvX\vXvXvXvX

ten

Lv.v.v.v.v.v.v.v.v
LV.V.V.V.V.V.V.V.V

1 '.
::;X : x : x : x : :
XvXvXv:

:-r:

igir-jOi

W jo!

n
<i

Ul

o

Q

o

CM

a
a

in

CI

a

a

LO
CM

O

a

CM

D

a

in

CM

a

a

Ul
CM

o
a

Ul
CM

a
o

Ul

CI

a
a

ui

CM

a

a

ul
CM

i—
'X

Q

CL

u

•x

LU

D

L_

I—

X

1 1
X
XI

LL

CM

h-
X
Q

X
UJ

X

CO

r-
X

X
Ul

X

ft

H
X

X
Ul

X

U"

u.
I

LL

Ul

X

r-
X

LL

'X
X

N

r-
X

X
Ul

X

CO

L-

X

LL

Ul

lL

X

X
X

i
-

xi
iOl

LL|

LU

1

1

[XvvXvX

V V.V.V. .v.v.v.v.v

L\v

!•: :■,::•:.

WmWM

r.v.v.v.*.*.*.*.*.*.*.*.*.*.*.

Lv.v.v.v.v.v.v.v.v

i .v.v.v.v.v

if! i^lO

? lull: 1 -'

til: : " :^

ice; ;o : fi
ijlV:U

•J)

a

T-

'T/

•JJ

N

'■£)

ul

^7

Cv

C*i

P"

U

UJ

h-

G

LU
LL
.-.

r-
<

L-

CO

CJ

z

H-
LU
-1
X

o

T—

+

to

CJ)

<

CJ

X

a

Q

a

CJ

D

a

10

CJ

a

D

Ul
CM

a

ui
CJ

a

Ul
CJ

X

Ul
CJ

d
ui

CM

a

Ul
CM

a

c

ID

CJ

1
1

1

1

J
OI

O!

ui!

C-!
1

I
j

Q

Q

.— .

1 : : x : x : x : xl
► v •:•:•:•: :•: :■:•:•:•:•:•:•: : :

XvX-XvXvXvXXv

L .V/.V.V.V

txcjxvx ■ :

fevX-MvlvlvX

!■-' ■■■■>

XvXv:;:;XvXv|i:

xvx^:vxvx»

: m &■

vXvM W
t-vx^x;-;^

tvxv» u:

! fevidfe

tvAvvX*

mm -M

ilit

l ft s

txv:«vXv>«
PP\$

a ' SO

3fi U Ul 1- 31

- :"-u.

o

T—

cm

I—

CJ

r-

CM

i—

CM

T—

CM

CM

r-

CM

CM

T—

T—

u

1 . i

r-

5

Lu

LL

r.

<

z

Q

u
Q
: 1

D

X

Lu
>

CM

i—

X

Ul

c

Q
rj

a

d

G

a
a

G

Q

G

G

Q

G

G
O

C

G

Q
Q

Q

X

G
□

,— %

i ;_ii

: LU : <[ :

Ijj- j- : x-

<! :•■! iQifhi

0: ; ; ;o.:p

DiZi

T—

r-

1—

T—

T~

T—

T-

7—

t—

T-

n

CD

o

O

LU
LL

CO

CJ

<

I—
ffl

n
Q

Q

O

n
G

Q
G

G

w

Q
G

G

■X

G

Q

a

G

1

! ; : => : ti. .1
iNi>#i<;

UJ u] Q

• :gc:a:; i

: ll. :

D

1™

CM

CO

■?

m

CO

N

CD

■Jl

cn

m

D
CO

O

a
a

a

a
a

O

d

a

a

a

O

Q

O
u

d

a

Q

G

1

Q!
d

WS ""Uu:

ije-UJ-

■ *.H • <. m ; ■ LL ; ^ ■
: i . i : ,~\ :

ibrH

:::::::
:::::::

i i i i i i i

cm

CO

■*

0)

f | 1

L fcXvXv*
tXvfcvXffi

b#v:vxi?
! Svx:x#
xv:v*x : :v:ffi
txvxxxxx-xjfc

i—

m

to

N

CD

a

m

-—
N
X
CO

CM

a

Q

d

a
a
d

a
d

a

a

1 1

d

a

a

d

LJ

a

Q

a

ol

G|

Ol

1

!
_|

1
1
!

1
1

1

1

-1

51
dl

!
1

\ XvXvxft
tXvXvX: v

fXvXvXvX-XvXv

XvX^pvX

I • •

IvX^X^XvX?
jvXvXvX

II

i

: xlvxx

• .:

:.»: MM

ifei

in; M i i

B
•_! ,-r I— IX

:z:Qi i :y 13
■":?' *in

q:uj; ;

:2

:£.:

: jf •

:•**.: : : :

z

1 —

a.

UJ

X

CM

i_
■X

r,

X
UJ

LL

CO

l_

CC

i i
X
LU

LL

DC

o

LL

lX

o

LL

LU

LL

LL

Q

LL
LU

CC

N

H

LL

o

a,

LL

■23

L-
LL

o

LL
LU
LU

CD

LL
i 1

i

LU
UL

CD

o

CL

LU

LL

N

Q

LU

<
S

H
CD
Ui

H

Co

x

X
'J

H

'-9

1 (

rj

X
Q

!

a
ci

n

a

n

5

Q

n

a

G

Q
Q

O

n

a

a
a

d

.-1

a

Q

G

45.00

40.00 -

IWnFDFKJnFWT PFRFAQHH PPHIFnT

two CDii^crc

/ \

/ \

35.00

i i

/ \

COSTS

30.00 -|-

?r nn

20.00 +

) /

I /
-u

n

LK^

Li

/ f\ \

/ / \\

^7

— r

\ \

n

= 0= [

15.00 --

10.00 4-

5.00 -r

0.00

-i 1 1 1

12 3

5 6

MONTH

10

PAGE 1

CnSTTRRC! WDRKSI-FET

»„■_■ ...... ...■....■.......■.■..........,.■■,.■..,...■■■........■..■.. .. . ■ . . . . . ...... ... . T— . . ...... ................ ..... ......■■■ ..1

GOSTTREM 'tfGRXSHEET
WtwOSPtKZS

DATE 25-MaHO

ORIGINAL COST= ; 20

1 L n J

! PREVIOUS i PROJECTED REMAINING ESTIMATED

ncruni ncruni rcruni : ururtic uurirLc nun t iric uuo i rt i

NUMBER DATE DATE i INTERVAL ; DATE ! TO COMPLETE ; COMPLETION

jkrr-ii it ,— <-,i .4 _j—-r^i ■* tkir*.i rr-

ir«rui luli "i^uLi iitu i

ECAC

*JNAL 10 20.00

EDRT1 10 1 10 3 20.00

:.jhi i. iii iu tj.uu

BRT3 3 2 1 10 7 20.00

•'hnr 4 j a 4 4 a ■-■ aa a a

E.3RT5 5 4 1 10 5 20.00

•HrtT .*• A C 4 4 A 4 AC A A

2ijni o o J i i u *t *.j.'j'j

■3RT7 7 6 1 10 3 20.00

■ir.T .*. n -> ■* 4 n a in r>rt
C-'P. 1 0(1 IU L iU.J'J

E0RT9 3 3 1 10 1 20.00

PhnT 4 A 4 A A 4 4 A A AA AA

turn i iu iu j i i ■ j u iU.uo

!
ill; : i 1

i i i i i I

7 3 3 10 1i 12

ESTIMATED COST PREDICTED j PREDICTED

COST SLOPE ADDITIONAL COST AT REPORT

OVERRUN OVERRUN COMPLETION NUMBER

6- ORi

r

3 COST 2(3x7] SUM 3 3/1 '"2 5x10 3+11

PCAC

i ; ; ; ■ ;

n

00 0.00 O.-QO 0.00 00 20.00 REPORT 1

*j

00 10.00 10.00 2.50 20.00 45.00 REPORT 2

Q

00 0.00 10.00 1.11 7.73 27.73 REPORTS

00 0.00 10.00 0.63 3.75 23.75 REPORT 4

n

tiCi n Qfl ifififi h ii'i ? nn n nn ccphpt r

5

00 10.00 20.00 0.56 2.22 27.22 REPORT 6

ij

00 0.00 20.00 0.41 1.22 21.22 REPORT 7

00 0.00 20.00 0.31 63 20.63 REPORTS

Q

00 0.00 20.00 0.25 0.25 20.25 REPORTS

o

00 0.00 20.00 0.20 0.00 20.00 REPORT 10

;

1

<

<

LU

Q_

■

n

■

U 1

1

1

0)

LU
|

h-
W

LU

CJ

SF LU

_l

2

n

l~!

i.i

~~

i-

H

2

uJ

LU

Q

h

T

[/]

n

u

111

z

C3

h

<£

X*

-l

>■

I

uu

o

- /

-1-

,— ',

LL

^x

CO

*~

_i

-1

h

H

/

^

,'

/
n— t

J I

n

rf

T m

+ CD

ij;,

4- ID

-=11-

D

a

a

a

in

CSJ

tn

h

CO

:

-\;

\\

! !

: :

1 !
j ;

i :
| :

G

1-1

LJ
r—
<
>

r-

ra

UJ

<

i—
CO

z
o

h"
Ul

CL

I

o

D
0.

n

LU

o

G

a

—a

o

G

ui

CI

a
a

in

CM

G

G

CM

c

CI

o
a

ui

CM

O

G

L0

CI

G

G

to

CJ

o

a

ui

CI

o
a
ui

CM

1- lT
rriuj;

niSi
£:£:

UJ : _J:
*X:Z :

I- i>J

iccjcc

£ LX
UJ:UJ

X LL

CO

1-
X

tii

X

•7

h-

X

X
LU

X

LO

X

X

UJ

X

g:n

<r- \-
iX IT
i

£:£

LULL!

LL^X

G

(X

o

LU
X

Xi

'-

X

LL.

X

X

L-l

xi
01

0.1

— i

i

:
:

!

1

i

i
1

LO

<

LU

EC

LU

1 —

til

-
LU
_l
X

Q
CJ

■_■

i—

Q
CJ

1

7

o
o

3)

GO

CC

m

rr

CO

CJ

T™

G

:U
:i ■ i

;h

n!0

; t- :lj

LU

:LL

k0

<5I=!t-

i j ' T : G
CJiQ

5

V.

t J • G • G
^ u u

'J G L0

a.; w;^

CO

Ifi

LO

CJ

X
CO

a

G

CO

CO

G:N

LO CO

did

CO : C 1

CJJ

1—

1 '-

CM

CD

X

X
CM

x;

CM j

1

;

;
I
:

1 1

f 1
i :

! :

o

j

X
I

CD

r—

ST

LU

H

o

LJ
"3
Q

E
LL

g

Hl

-i
0-

>

u

LU
*-
<

U

— .

a.

o

T—

a

i—

o

T—

a

r-

G

T—

o

1—

a

r-

o

r-

a

r-

a

:Cl

i ■ i

!i-

It-ICJ

:uj
III

.,> —

■"5

_j:

■Jiz:

Q i :
<

': -. : G

5 °

L0

ui

—

Ifl
w

T—

r,
G

iP-33

LO : CO

'X

X
X

G .

x!

x j

i

i j

i !

i ;

1 i

CO

LU

H

<

Q

LL.

D

<
'•v.

>•

CC
H

1-

r-

T-

P"

r—

T—

i—

i—

r™

I—

J^PQ:co!9r: j<

' <r r- ' J ^

>44JJ_

iOiG
iu'LO

G M
j x'

CM

CJ

CI

CO

f

X

T—

'X' : C'J

co:c-j

T ~ : »™

lT»

X

T—

X
X

X

[5 (
O i

t

1 i

i

1 I

i y

Uj:
1 XI

1 B-

cm

CO

D

LU

X

u-

LL

X

LU

CC

LU

H
<

a

r-

CM

CO

<*

Ifl

to

N

G

'T;

:/r.

H 1

: '_i : ~.

n =

X
Q

C-J

G

a

,— i

CO

X

G-G

.— ,— '

LO : CO

G

Q

,— i

X

Q
— i
G

i

G i

— j 1

55!

i " :

i a;
Si

P.

—
CM

1—

H

X

Q

LL
I > I

iX

LU
r~

r-

<M

CO

7

LO

(0

K

CD

■Tt

a

T—

! ;Ci

:_:n
:0:.=5

:o:'-;
jo!"

G

X
G

G

G

UI

G

QjO

'—1 ; 'J

G j G

Q

a

u

Gi

1 K :
CO;

0-

V

n

II

1-

33

n

r

_i
z

o

c

LL

LU

LL

cc

Ixi

3

-

T
L_

1-
\L

1 1

0.

LU

lX

CM

j

CC

CL

LU
X

CO

X

o

LL

LU

■7

I—

ir

Q

LL

UJ

X

LO

i—
CC

X

LU
X

l_
LL

Lu

X

■x

X

Lu

X

CO

X

X
uJ

X

G)

t —
X

.

LL

LU

iX

G

r"

l_

ir

X
X

i^^ g ■'-'

j N :3:»:(I:g

:£:V;U lX
• |- i '--■ ■ > i Q
■CO; :U: ,

;ll!: ;";i

:n : i-i
i Q ; a
: O i LO

X
LO

Q
LO

n

a
ui

q i a
if] 1 ui

n

a
ui

ui

G

Q

.-!
iiJ

i

. 1

'

^

-.

CO

j_

CO
<

•jj

£E
n

h

m

u
o

IX!

(3
CE

4

' 4

* --»

5

o N

s

3

o

~~v

<

^

Q.

ID

1

I

a

-4

a

— 5)

CD

-t- CD ■->'

u

I
ID K

o

- 1 - ^r

en

CO

n

ffl

Q

UJ
H
<
2:
H
CO
UJ

<

m

U

z

H
LU

LL
5

D
0.

z

y

G
G

d
CM

O
O

ui
eg

o

o

ui
CM

O

G
111
CM

o
o

ui
CJ

o

G

ui

CM

o

o

in

CM

G
O

ID

CJ

G

G

ui
CM

G
G

ui

CJ

1-

G

n

G
UJ

G

G
G
G

Z

It"
lr-
|CC

i ;0

: u-

LU
CC

CJ
H
G

G
G
G

G
h-
G

u.

LU

G

G

G
G
G

m

H

G

G
G
G

(OiNIGIC!)

G G G G

G : G : G : G
LU-UJ ; LU UJ

iriccG

a

r—

H
C

0-
g|

'1
1

in

<

UJ

CC

UJ

111

H
LU
_J
n

T-

Q

!—

_1

n

I

•*

n

0)

G

N

T—

G

Ifi

1

G

CM

T—

G

OJ

LJ

UJ

H

G

LU

LL

0.

<

t-
G

z

H
U
_i
G

.- DO

■\--0-6
(Oia.iC'J

G
G

ui

CO

CM

G

in

N

G

G

LJ

G

G

U0:Gii-:G
Q:CD N Ui

CO Oi CM ^

a

G

in

Q

<r>
1

a
<
1

CO

■?

Q
u

H

u

UJ

DC

LL

Z

K
LU
_J
Q.
g

UJ
<

D
G

Z

o

I—

o

i—

o

C'J

O

T-

O

1
/

O

O

T—

G

G

O

T—
1—

u
UJ

t-

o
5

G
G
n

<

Z

t-
Q
Q
<

z

3
cc

G
G
>

^: :n

X; ;h

■ i i

G
G

G
CM

/

G
N

N

G

m

N

i—

r-

G
Q

G

G N G 3)
m • (0 ■ t- ; CS!

ui g (Si o

o

Q

G

t

UJ

<
Q

m

UJ

H

<

Q
a.

D

<
>

iX

UJ

z

r-

r-

1-

I—

T~

T~

t—

T—

I-

I—

G

T—

>-
CO

CJ

u

g

_i

CMi L

< i :g
^: id

ffl:

G

in

CM

CM

CJ

CM

G
G

G

G

T—

(T) t i G 3)

g:cm:o:g

r~ • T" • r* : O

t — r

G

COST TREND WORKSHEET

CM

CO

_)

>

u
cc

a.

i-

G

o

a.

Ul

LL

UJ

1—

<

Q

o

t-

0J

g

•3"

UD

G

N

G

G

G

2 o

Di :d

CO:

G
n

G

T—

G

Q

a

C'J

'— 1

Q

d

G

G

Q

G

O ; O ; O ■ O

Q : G'Q : G

d:d:d:d

lfi:G;N;CD

g!

gI
1

a

CM

-

t-

cc

LL

LLi
CC

LU
i-
<

Q

-

CM

CO

t

W

G

K

G

G

G

G

% jo
S. d

C>J

: D

a

G

G
—

a

G

G

i—

G
Q

G

a

G
—

o'a g g

O : Q : O • G

o o o o

a

o

o

I—

!

■

9

r—

CO

-I

<

z

'-■

a:

cc

LL
i . i

fcc

at

LU

UJ

Z

1-

G

a.

LU

LL

CM

H
IX

Q

G

UJ

cc

CO

H

G

o

LL
LU

at

t

t-
cc

G
LU

at

Ifl

H

g

LL
LU
LL

G
\~

cc

a.

Ul

LL

u-

at

a.

LU

at

CD

h-
G

a.

LU

G

t-
G

a.

LU

G

G

t-
G

a.

LU

G

N

LU

1-

'0'

t-

H

CO
UJ

Cu

z

-J

G
LU
->

b

t-i
fjj:

0; i

« n

C n

?: id
u

li 1

CD]

o
ui

-

5
ui

I - !

G
ui

a
a
in"

0:G!G;G

idjirijiri u-)

n

a
in

-

APPENDIX D

*_-«-'0 I flft'--.Al_l v l

i it„„in.-.~~-,4<
jicjirvcuutu i

2 !=OIALOG.BOX(BOX)

3 |= iF(A2=FALSE,G0T0 #REFJ)

5 I =OPEN rCOSTTR.XLS" 1)

.-»r\r- hi /ii#-v*-*j-v i i r^r~ \/i #-ji -i*

='.'Hc in UOa i i K h.XLi ' ,\f

10

I I | =A0 I IV A \£\ " UU5I IKALS }

12

-

13

14 i=D IALOG. BOX ( MENU)

ir i=!F/!i?=i nnTn/iwrnATFU

1b i=IF(M 2=2.GQ I Q(U PDA I E))

17 i=i F(H2=3.GOTO(preport))

1 fl ! = JF( ! 1 2=4 ,GOTO (PC HART))

19 !=1F(I12=5.G0T0(EX1T))

on - !

1

21 i

22 !=1NPUT ("INPUT TH E PROJEC T NAM_E"Z'TO(^E^:TJITLE7

23 ' =SELECt / "R1 1C3:R2 1 07'^

24 |=CLE/

25 =S ELECT("R30C1:R33CS")

\

26 =C L£AR(3) j

I 27 j=INRrr^lNRJTTHE ESTIMAT E COS T OF THE CO NTRACPM, "TOTAL ESTIMATE COST 1 )
! oq LiKjpi rrf'iMP! rr • FMTTH ^f Don 'FT 'M MPMTHO" i "I cwniTj dp Don ir:r*T"\~

| i.u ; ~n ii *.* i ^ it ii *j I l-1_i tv3 i ii vi i i i»-tit_»_' 1 hi rivni i iu , i , i—i—i rjiilvi i i tvui_v i /

j 29 j =S ELECT (!C2)"" j

30 j=FORM ULA(A 22)
qi r

32 |=SbU:C r(JB4)
j3 |=FORMULA(A27)

34 !=SFI FCTflF1?>

■ ;-_-—--■ • v 1.-

_35_j=FORMULAiA28)

37 !=INPUTf"ENTERTHE REPORT DATE". 1. "REPORT DATE"}

33 |=INPUT ("INPUT THE ESTIMATED COST AT COMPLETION", 1 /'ESTIMATED COST AT COMPLETIG

39 j=iNpyTHNPUT THE PROJECTED COMPLETION TIME FRAME", 1 ."PROJECTED COMPLETION d!

40 'REPORT DATE

41 j=SELECT(!B12)

■ -|

4? i=iFfGFT OFI I ffiW'>A37 SFI FCTf"Rr+1im GOTO/PRFVIOUKU

43 =GOTO (A42)

44 ! PREVIOUS

45 i =SELECT("R[-1iq+11")

46 | =IF(GET. CELL(8)Q3,SELECTC'R[- 11C'1GQTQ(FQUND)

47 pGOTO(A46)

40 IhUUNU

Page i

i

An

I— OL.U.U i \ r\^[-ij y

!
I

50

l=GET.CELL(5)

i

I

I si

jiNSERT ALL VALUES

I

1! co

l-6pi POT/inio\ .

1
1

53 l=IF(G ETCEL L(5 )<>A37,SELECT ("R[-t-1]C"),GOTO(INSERT ))

a4 |=(.ici.OcLL^ b)

! 55 !=GOTO(A53)

56

58 i=SELECT("RC[+ 1j")
55 j=ALiGNMENT(3)

SO = FORMUU (A50*

61 1=SELECT("RC[+1]")

62 I = IF(A37=1, FORMU LA (1),FQRMULA(A54- A50))

63 j= SELECf("RC[*l]")

J

-l

64 j=r ORMULA(A39)

S5 i=S ELECT("RC[+1]")

66 | =FORMULA ("=RC[- 1 ] -RC[-4] ")

I 67 j

l cq !-Ofri irr*T/"Df~*rj.n"\

89 !=FORM UL A(A38)

70

71 !=sfi FnTrpr+iRirr-Ki"\

+-r

! 72 l=GEJ.CELL(5J

p3 |=FORMULA(A38-A72)

74 l=SELECTf'RCr+n"\

75

=FQRMUU("=RC[-1]*R[-18]C[+21»2")

7S

77

78 |=3ELECT( "RC[+ 1]")

7Q i=FnPMi ii A/"=nnr-i-unr-nr^

• " I •_r. , .r..jir-^i. , j : :~i,_u • *i ' J- /

80 | i

1

ui i — oi_i_i_v^ i ^ rv-_<[+ i j i

i rv i=FORMUi A/"=Rr:r-n,'fmr-imnr-?ir?VT

-| ! ■ ■■■ y ■ - t -J'.U" 'k " ~J--A-JZX1 =i_X_

83

QA

85 l=SELECTf'RCr+in

-r-

86 j= FQRMULA(" =R[-18]C[+1]»RC[ - V^_

R7 I

88 =3ELECT("RC[+13")

|=FORMULA("=R[-18]C[-»-1]+RC[-1]")

90 bAOTJVATFfnOSTTPE XI C:"\

51 i =CALCULA i E.NOWQ

I 95 {preport

hjjl ' ' 1

I 31 I

9b i=ShLhCl(!A1:G39)

Page 2

m Loct nniiiiT incin
01 | — PL. i .rniM i .HnLn^

98 j=PRINT(1,.,1,FALSE,FALSE.1)

3a

1 fin Ur:nTn.'U*lMUPkll l\

tuu i — *.4 »^ i »«* i i ■ i n 1 1 1 1 ■ 1 1— i < *_'y

101

102!

1Q3!

|04 print chart

inc I

106

|107i __!

iTOSi = ACTIVA TE ("C OS ffRE.XLC' ^

109;

|110| =ATTACH.TEXT(1)

ill !=FORMULA'A22^

1 1Z i=HKlN I {i„J,l-ALSt,l-ALSt.i)

IIOLl r^TWI JTC/lir-nOTTn VI 0"\

[ i i j j — M'^- 1 i f rt 1 i_^ Owo i i rxAi-o /

114

h— — i

| Il5j

1 1 1 c i -VirrriS/M a ikimcmi n

; i i u ) — *_•!»_' i »_' ii ■ ir»n ii - it_i t »_'/

11171

iTTVl — 1

I IQI

1 11 q I

l____l_\ i

I122 J

1 123 1

\lOA i-bfr-fi IDMA

( I A-T j — I U_ I »_' I U 1 W

i125 !

__ ne I

Page 3

i — i — i — r

|£0

-

33

01

*

co ! ro

i

- 1

— ,

B

CO Kl

j

ca

■-"

"

U |»H

i i I i

iro
i

i

i
!

1

(\-

TO

ro

f j

&

m
r~

m

H

n

>

r

r 3

cn

01 : Ol '•■

i I

Ml

■ill

!
; 1 1

1 ! !

: 1

i i

!
ill:
! 1 I i

ro

CO

CO

i j

— L

CO

03

£33

00

-j.:o
03;

CO

_L

-•J
"--i

'10

'CO

CO:CO: !

j ! !

: : 1

!

i i i 1
1 1

jo;

! 1
1 i

: ; i

: : I 1

: : 1 i

CO

'-

CO

CO

CO

01

en

---j

CO

CD

to i o

CO

"--J

4^

CO

CO : — *■ :

ojroj i

i i

; 1 !

1 i !

; j 1

i i i ,M i

! ! I
: i i

iil]

■ ■ 1 !
: : 1 i

; i i

(33

CO

CO

re
D

CO

ro

s

(33

i !

: j 1

i i I 1

1 i j

■ ill

1 — 1

i j ' ! i

: i
j

i

Mil
i I

CD

CD

CO

i ! 1 i

i i !

i i
: i !

: I !

MM

1 i

: : 1 1

MM

M 1

= 1 1

! ! i 1 1 ! !

Li_iLj._L_i._i.Li

■ i [ i i I :

!

s i !

CO CD p.) 03 CH 4- j CO

! i 1 i ! 1

| 1

-

mim

T!

i_

,

- .-, n

ti j n ■ r

!_?

T

X

u

Zi

Tl

_;

: r^

a) -r

c

Tl i r\ ■ i

3

H

_ .

2

n

H

: Jy

_■• . '

TJ

m-X. i

.-

7j

-.

—

_j

i>

1 i

I:- . _i :

ji

___

1

H

H

; I~

1_

r

Hi"0; i

i

>

o

i/J

m

m

m

miC: I 1

( i

X

~ j

Tl

JJ

T

,

D!_l:

I)

>

X

Tl

m

: t —

— ■

midi ! !

o

13

11

TJ

i —

^_

< u. 1

H

H

T

m

m
-i

Tl

H

( )

3

-1

CO

,• : 2- : 1 i

tjW j |

CD:- 1 :

> 5 1

i

1

m ■ ! -- ; ■

i

_J . -J :

|

' "7" '

-

i
1

:Xi

1

m:

: — ■

1

-i:

I!

: w I

j

I

:_.:

i

!_jj

i

■ f"" 1 - ■

i

. ! j

■

i :

■
i i

i
i

i

i

j

i

i
1

1

i

!
i
i

i
i

fc/V-5^

Thesis
B16863 Baker

c.l Completion cost trend
analysis. '

```