INTERNATIONAL JOURNAL OF

PROJECT
MANAGEMENT
International Journal of Project Management 24 (2006) 289–302
www.elsevier.com/locate/ijproman

A comparison of different project duration forecasting methods

using earned value metrics
a,1 b,c,*
Stephan Vandevoorde , Mario Vanhoucke
a
Fabricom Airport Systems, Rue Gatti de Gamond 254, B-1180 Brussels, Belgium
b
Department of Management Information, Operations Management and Technology Policy, Faculty of Economics and
Business Administration, Ghent University, Hoveniersberg 24, B-9000 Gent, Belgium
c
Operations and Technology Management Centre, Vlerick Leuven Gent Management School, Reep 1, B-9000 Gent, Belgium

Received 28 June 2005; received in revised form 13 September 2005; accepted 14 October 2005

Abstract

Earned value project management is a well-known management system that integrates cost, schedule and technical performance. It
allows the calculation of cost and schedule variances and performance indices and forecasts of project cost and schedule duration. The
earned value method provides early indications of project performance to highlight the need for eventual corrective action.
Earned value management was originally developed for cost management and has not widely been used for forecasting project dura-
tion. However, recent research trends show an increase of interest to use performance indicators for predicting total project duration. In
this paper, we give an overview of the state-of-the-art knowledge for this new research trend to bring clarity in the often confusing
terminology.
The purpose of this paper is 3-fold. First, we compare the classic earned value performance indicators SV and SPI with the newly
developed earned schedule performance indicators SV(t) and SPI(t). Next, we present a generic schedule forecasting formula applicable
in different project situations and compare the three methods from literature to forecast total project duration. Finally, we illustrate the
use of each method on a simple one activity example project and on real-life project data.
 2005 Elsevier Ltd and IPMA. All rights reserved.

Keywords: Earned value; Earned duration; Earned schedule; Project duration forecasting

1. Schedule performance indicators project and it helps to evaluate and control project risk
by measuring project progress in monetary terms. The
Earned Value Management (EVM) is a methodology basic principles and the use in practice have been com-
used to measure and communicate the real physical pro- prehensively described in many sources (for an overview,
gress of a project and to integrate the three critical ele- see, e.g. [1] or [2]).
ments of project management (scope, time and cost Although EVM has been setup to follow-up both time
management). It takes into account the work complete, and cost, the majority of the research has been focused
the time taken and the costs incurred to complete the on the cost aspect (see, e.g. the paper written by Fleming
and Koppelman [3] who discuss earned value management
from a price-tag point-of-view). Nevertheless, earned value
*
Corresponding author. Tel.: +32 9 2643569/2109781; fax: +32 9 management provides two well-known schedule perfor-
2644286/2109803. mance indices, the schedule variance (SV) and the schedule
E-mail addresses: stephan.vandevoorde@fabricom-gti.com (S. Vande-
voorde), mario.vanhoucke@ugent.be, mario.vanhoucke@vlerick.be (M.
performance index (SPI), to measure project progress. The
Vanhoucke). SV is the difference between the earned value (EV) and the
1
Tel.: + 32 0 2 3703111; fax: + 32 0 2 3703222. planned value (PV), i.e. SV = EV  PV (for a graphical

0263-7863/$30.00  2005 Elsevier Ltd and IPMA. All rights reserved.

doi:10.1016/j.ijproman.2005.10.004
290 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

In order to overcome the anomalies with the earned

value schedule performance indicators, Lipke [4] intro-
duced the concept of earned schedule (ES). In this method,
the earned value at a certain (review) point in time is traced
forwards or backwards to the performance baseline (S-
curve) or PV. This intersection point is moved downwards
on the X-axis (the time scale) to calculate the earned sche-
dule ES (see Fig. 1). Hence, the ES is found by identifying
in which time increment of PV the EV occurs. It translates
the EV into time increments and measures the real project
performance in comparison to its expected time perfor-
mance. The corresponding schedule performance metrics
are:

SVðtÞ ¼ ES  AT; ð1Þ

SPIðtÞ ¼ ES=AT; ð2Þ
Fig. 1. SV versus SV(t). where AT is used to refer to the Actual Time.
In contrast to the SV, the SV(t) is expressed in time
presentation, see Fig. 1). Note that the PV is often denoted units, which makes it easier to interpret. A SV(t) < 0
as the BCWS (Budgeted Cost for Work Scheduled) and the (>0) indicates the number of time units that the project
EV as the BCWP (Budgeted Cost Work Performed). The lags (is ahead of) its expected performance. The behav-
SV measures a volume of work done (i.e. earned) versus iour of SV(t) over time results in a final SV(t) that equals
a volume of work planned. However, the SV does not mea- exactly the real time difference at completion (while the
sure time but is expressed in a monetary unit. If SV < 0, a SV always ends at zero). The same holds for the SPI(t)
lower volume of work has been earned as planned, and the indicator, which has a final value reflecting the final
work is behind plan. If SV > 0, a higher volume of work project schedule performance (while the SPI always
has been earned as planned, and the work is ahead of plan. equals 1).
If SV = 0, the earned work is exactly as planned. At the
end of a project, the EV = PV = BAC (budget at comple- 2. A generic project duration forecasting formula
tion), and hence, the SV always equals 0. The SPI is the
ratio between the earned value and the planned value, i.e. Earned value metrics have been widely used to monitor
SPI = EV/PV, and is a dimensionless indicator to measure the status of a project and forecast the future performance,
the efficiency of the work. If SPI < 1 (=1, >1), the schedule both in terms of time and cost. The use of the metrics to
efficiency is lower than (equal to, higher than) planned. At forecast a projectÕs final cost is numerous and is outside
the end of a project, the SPI is always equal to 1. the scope of this paper (for an overview, see, e.g. Christen-
The interpretation and the behaviour of the earned sen [5] who reviews different cost forecasting formulas and
value management performance indicators SV and SPI examines their accuracy). In this section, we elaborate on
over time have been criticized by different authors [4]. First, the use of the metrics to forecast a projectÕs final duration
the SV is measured in monetary units and not in time units, by different methods. A generic project duration forecast-
which makes it difficult to understand and is therefore ing formula is given by:
often a source of misinterpretations. Secondly, a SV = 0
EACðtÞ ¼ AD þ PDWR; ð3Þ
(or SPI = 1) could mean that a task is completed, but could
also mean that the task is running according to plan. where EAC(t) is the estimated duration at completion, AD
Thirdly, towards the end of the project, the SV always con- the actual duration and PDWR is the planned duration of
verges to 0 indicating a perfect performance even if the work remaining.
project is late. Similarly, the SPI always converges to 1 We use the EAC(t) metric to refer to any forecasting
towards the end of the project, indicating a 100% schedule metric for a projectÕs total duration (note that the abbre-
efficiency even in the project is late. As a result, at a certain viation EAC – without (t) – is usually used in the liter-
point in time the SV and the SPI become unreliable ature to refer to the cost estimate at completion) and the
indicators, and this ‘‘grey time area’’ where these indicators AD metric to refer to the actual duration of the project
loose their predictive ability usually occurs over the last at the current time instance. Moreover, the PDWR met-
third of the project (expressed in percentage completion, ric is the component that has to be estimated, and heav-
see [4]). However, this is often the most critical period ily depends on the specific characteristics and the current
where the forecasts need to be accurate, since upper status of the project [1]. In Table 1, we distinguish
management wants to know when they can move up to between six different project situations based on the clas-
the next project stage. sification described in [1].
 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302 291

Table 1
The estimated PDWR depending on the project situation (based on Anbari [1])
Situation Forecasting method Comments
Anbari [1] Jacob [6] Lipke [4]
EAC(t) as originally planned Monitor schedule The final project duration will be as planned,
regardless of the past performance. This situation
may be dangerous, as unattended problems mostly
do not resolve themselves (‘‘weÕll catch up during
the commissioning phase’’)
PDWR is new Re-schedule The original project assumptions are no longer
valid for the remaining work (due to changed
conditions). The use of performance indices to
predict is obsolete and a new schedule for the
remaining work needs to be developed
PDWR is very high Re-schedule Quality problems are irreversible and a lot of extra
time is needed to fix the problems (occurs mostly
in the late project stage). Stakeholders usually loose
their interest in the project (‘‘If this project ever
finishes, it would be a miracle’’)
PDWR according to plan EAC(t)PV1 EAC(t)ED1 EAC(t)ES1 Past performance is not a good predictor of future
performance. Problems/opportunities of the the past
will not affect the future, and the remaining work
will be done according to plan
PDWR will follow current EAC(t)PV2 EAC(t)ED2 EAC(t)ES2 Past performance is a good predictor of future
SPI trend performance (realistic!). Problems/opportunities of
the past will affect future performance, and the
remaining work will be corrected for the observed
efficiencies or inefficiences
PDWR will follow current EAC(t)PV3 EAC(t)ED3 EAC(t)ES3 Past cost and schedule problems are good indicators
SCI trend for future performance (i.e., cost and schedule
management are inseparable). The SCI = SPI * CPI
(schedule cost ratio) is often called the
critical ratio index
Planned value rate Earned duration Earned schedule

The first project situation assumes ideal circumstances displays the terminology used to refer to Eq. (3). The col-
and does not require any forecasting since the project is umn labelled with ‘‘Assessment indicators’’ displays the
considered to be on plan. The second and third rows refer terminology used to measure the additional effort needed
to project situations where forecasting (i.e. estimating the to finish the project within the project deadline. The specific
PDWR) is useless due to the changing conditions or irre- calculation of these metrics will be explained in detail in the
versible problems. Hence, in the remainder of our paper, following three sub-sections.
we will focus on the last three possible project scenarios.
In these cases, we assume that the PDWR is according to 2.1. The planned value method
plan (scenario 4), follows the current SPI trend (scenario
5) or follows the current SCI trend (scenario 6). Each fore- The planned value method is described by Anbari [1]
casting technique described in the three following sub-sec- and relies on the planned value rate which is equal to the
tions will be discussed from these last three project average planned value per time period, i.e. PVRate =
scenarios point-of-view. BAC/PD where BAC is used to denote the budget at com-
To the best of our knowledge, only three project dura- pletion and PD to denote total planned project duration.
tion forecasting methods have been presented in literature. This method assumes that the schedule variance can be
In the remainder of this paper, we refer to these methods as translated into time units by dividing the schedule variance
the planned value method [1], the earned duration method [6] by the planned value rate, resulting in the time variance TV
and the earned schedule method ([4], and further developed as follows:
by [7–10]). However, the many notations, abbreviations
and often confusing metrics used to describe these three TV ¼ SV=PVRate ¼ ðSV  PDÞ=BAC
methods unnecessarily complicate the comparability of ¼ ðEV  PVÞ  PD=BAC. ð4Þ
these methods. In order to shed light on the confusing ter-
minology, we summarize the overwhelming amount of syn- According to the project characteristics (reflected by the
onyms taken from these various literature sources in Table last three situations of Table 1), the following forecasting
2. The column labelled with ‘‘At Completion indicators’’ formulas have been derived:
292 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

Table 2
Terminology used in comparison papers under study
Baseline Anbari [1] Jacob [6] Lipkea
SAC Schedule at completion PD Planned duration PD Planned duration
Status of the PVRate Planned value rate ED Earned duration ES Earned schedule
project AT Actual time AD Actual duration AT Actual time
SPI Schedule performance SPI Schedule performance SPI(t) Schedule performance
index index index time
SV Schedule variance SV Schedule variance SV(t) Schedule variance
time
TV Time variance – – –
CR Critical ratio – – SCI(t) Critical ratio time

TEAC = AT + TETC EDAC = AD + UDR EAC(t) = AT + PDWR

At completion TETC Time estimate UDR Unearned duration PDWR Planned duration
indicator to complete remaining for work remaining
TEAC Time estimate EDAC Estimate of duration EAC(t) Estimate at completion time
at completion at completion

– – – – IEAC(t) = AT + PDWR/PF
– – – – IEAC(t) Independent estimate
at completion time

Assessment – – TCSPI To complete schedule SPI(t) to gob To complete schedule

indicator performance index performance index for PD
– – – – To complete SPI(t)c To complete schedule performance
index for latest revised schedule (LRS)
a
The terminology used is based on the presentation by Lipke and Henderson ‘‘Earned schedule – an emerging practice’’ presented at the 16th Annual
International Integrated Program Management Conference, November 15–17, Virginia.
b
The SPI(t) to go is equal to the TCSPI (Eq. (15)) or the TCSPI(t) (Eq. (22)) of the current paper.
c
The to complete SPI(t) equals the TCSPI-LRS (Eq. (16)) or the TCSPI(t)-LRS (Eq. (23)) of the current paper.

 Duration of remaining work as planned The performance factor is used to adapt the future perfor-
mance to the past performance (depending on the project
EACðtÞPV1 ¼ PD  TV. ð5Þ characteristics) and reflects the last three situations of
Table 1, as:
 Duration of remaining work follows the current SPI
trend  PF = 1: Duration of remaining work as planned
EACðtÞPV2 ¼ PD=SPI. ð6Þ
EACðtÞED1 ¼ AD þ ðPD  EDÞ=1
 Duration of remaining work follows the current SCI ¼ PD þ AD  ð1  SPIÞ. ð9Þ
trend
 PF = SPI: Duration of remaining work with SPI trend
EACðtÞPV3 ¼ PD=SCI. ð7Þ
EACðtÞED2 ¼ AD þ ðPD  EDÞ=SPI ¼ PD=SPI. ð10Þ
Note that the terminology of Anbari [1] is somewhat differ-
ent since he proposes the Time Estimate at Completion
 PF = SCI: Duration of remaining work with SCI trend
(TEAC) and the Time Estimate to Complete (TETC) to re-
(note that this formula is not given in [6])
fer to the EAC(t) and the PDWR (see Table 2).
EACðtÞED3 ¼ AD þ ðPD  EDÞ=SCI
2.2. The earned duration method ¼ PD=SCI þ AD  ð1  1=CPIÞ. ð11Þ
The earned duration method is described by Jacob [6] In situations where the project duration exceeds the
and extended by Jacob and Kane [11]. The earned duration planned duration, and the work is not yet completed, the
ED is the product of the actual duration AD and the sche- PD will be substituted by the AD in the above mentioned
dule performance index SPI, i.e. ED = AD * SPI, and formulas. In these cases, the formulas are:
hence, the generic ‘‘earned duration’’ forecasting formula
is: EACðtÞED1 ¼ AD þ ðAD  EDÞ=1 ¼ AD  ð2  SPIÞ;
EACðtÞED ¼ AD þ ðPD  EDÞ=PF. ð8Þ ð12Þ
 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302 293

index for LRS

EACðtÞED2 ¼ AD þ ðAD  EDÞ=SPI ¼ AD=SPI; ð13Þ

index for PD
completion PF = SCI(t)b
completion PF = SPI(t)
Estimate of duration at

Estimate of duration at

Estimate of duration at
EACðtÞED3 ¼ AD þ ðAD  EDÞ=SCI

schedule

schedule
completion PF = 1
¼ AD  ð1  1=CPI þ 1=SCIÞ. ð14Þ
An additional assessment metric given by Jacob [6]

To complete

To complete
performance

performance
measures the additional effort needed to finish the project
within the project deadline. This corrective action metric

EAC(t) = AD + PDWR/PF
related to the schedule performance is called the ‘‘To

Earned schedule method

Complete Schedule Performance Index’’ (TCSPI) and is
calculated as:
TCSPI ¼ ðPD  EDÞ=ðPD  ADÞ ð15Þ

TCSPI(t)-LRS
EAC(t)ES1

EAC(t)ES2

EAC(t)ES3
or

TCSPI(t)
TCSPI-LRS ¼ ðPD  EDÞ=ðLRS  ADÞ. ð16Þ

The former measures the additional effort needed to finish

the project within the planned duration while the latter
measures the effort to finish the project with the latest re-

index for LRS

index for PD
vised schedule (LRS) duration.

Estimate of duration at

Estimate of duration at

Estimate of duration at
completion PF = SCIa

schedule
schedule
completion PF = SPI
2.3. The earned schedule method

completion PF = 1

To complete
To complete
performance

performance
The earned schedule method to forecast project dura-
tion has been recently introduced by Henderson [8], and
is an extension of the work done by Lipke [4]. Henderson
EAC(t) = AD + PDWR/PF
has illustrated the validity of the earned schedule concept
Earned duration method

by applying it on a portfolio of six projects [7] and on a

small scale but time critical information technology soft-
ware development project [9]. The earned schedule ES

This forecasting formula does not appear in Jacob [6], and has been added by the authors.
This forecasting formula does not appear in Lipke [4], and has been added by the authors.
TCSPI-LRS
can be mathematically expressed as:
EAC(t)ED1

EAC(t)ED2

EAC(t)ED3

TCSPI
ES ¼ N þ ðEV  PVN Þ=ðPVN þ1  PVN Þ; ð17Þ

where N is the time increment of the PV that is less than

current PV, PVN the planned value at time N and PVN + 1
is the planned value at time N + 1.
The generic earned schedule duration forecasting for-
Estimate of duration at

Estimate of duration at

Estimate of duration at
completion PF = SCI
completion PF = SPI

mula is:
completion PF = 1

EACðtÞES ¼ AD þ ðPD  ESÞ=PF. ð18Þ

The performance factor used depends on the project

EAC(t) = AD + PDWR/PF

situation:
–

–
Planned value method

 PF = 1: Duration of remaining work as planned

Terminology used in the current paper

EACðtÞES1 ¼ AD þ ðPD  ESÞ=1

EAC(t)PV1

EAC(t)PV2

EAC(t)PV3

¼ AD þ ðPD  ESÞ. ð19Þ

 PF = SPI(t): Duration of remaining work with SPI(t)

–

trend

EACðtÞES2 ¼ AD þ ðPD  ESÞ=SPIðtÞ. ð20Þ

At completion
indicators

Indicator
Assessment

 PF = SCI(t): Duration of remaining work with SCI(t)

Table 3

trend (note that this formula is not given in any of the

earned schedule papers)
b
a
294 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

EACðtÞES3 ¼ AD þ ðPD  ESÞ=ðCPI  SPIðtÞÞ The details are given in Fig. 2, which reveals that, at the
third week reporting period (W3), the project is overspent
¼ AD þ ðPD  ESÞ=SCIðtÞ. ð21Þ
and delayed. We consider two situations: linear and non-
The ‘‘To Complete Schedule Performance Index’’ or linear planned values.
TCSPI(t) can be calculated as: The total project duration for the linear case can be esti-
mated by means of the three forecasting methods as fol-
TCSPIðtÞ ¼ ðPD  ESÞ=ðPD  ADÞ ð22Þ lows. The Planned value method calculates the planned
or value rate as PVRate = BAC/PD = € 35.000/7 weeks = €
5.000/week and consequently, the time varience
TCSPIðtÞ-LRS ¼ ðPD  ESÞ=ðLRS  ADÞ ð23Þ
TV = SV/PVRate = (€ 12.000  € 15.000)/€ 5.000/
and measures the additional effort to finish the project week =  0.6 weeks. The Earned duration method relies
within the planned duration or the revised duration, on the earned duration of week 3 that is equal to
respectively. Remark that the TCSPI(t) and the ED = AD * SPI = AT * SPI = 3 · 0.8 = 2.4 weeks. The
TCSPI(t)-LRS is denoted as the ÔSPI(t) to goÕ and the Ôto performance needed to finish within the planned duration
complete SPI(t)Õ in Table 2. The terminology used through- is TCSPI = (PD  ED)/(PD  AD) = (7  2.4)/(7  3)
out this section has been summarized in Table 3. = 1.15, denoting that for each time unit that we spend on
the remaining work, 1.15 time units need to be earned in
3. Forecasting duration examples order to finish on plan. The Earned schedule method calcu-
lates the earned schedule as ES = N + (EV  PVN)/
In this section, we illustrate the use of the three forecast- (PVN + 1  PVN) = 2 + (12.000  10.000)/(15.000  10.000)
ing methods on a single-activity project example with and = 2.4 weeks and consequently, SV(t) = ES  AT =
without the presence of a learning curve. Moreover, we 2.4  3 = 0.6 weeks and SPI(t) = ES/AT = 2.4/3 = 0.80.
apply these forecasting methods on data at a higher The performance needed to finish within the planned
WBS-level of three real-life projects from Fabricom Air- duration equals TCSPI(t) = (PD  ES)/(PD  AT) = (7 
port Systems, Belgium. 2.4)/(7  3) = 1.15. Table 4 shows a summary of the fore-
casted project duration results based on the previously
3.1. Forecasting at activity level calculated measures and the values for the assessment indi-
cators by each method. All forecasting methods yield sim-
In order to illustrate the three forecasting methods, we ilar results, regardless of the method used, except the ED
display the EV metrics of a single activity example for method with a continuing SCI trend. This has also been
the installation of TFT monitors (Thin Film Transistor). observed by Jacob and Kane [11], who attribute the

Fig. 2. Earned value metrics on the activity level with (right) and without (left) a learning curve.
 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302 295

Table 4
Forecasted duration (PDWR) and the corresponding assessment indicators (TCSPI and TCSPI(t)) for our example activity project
Case Linear PV Non-linear PV
Anbari Jacob Lipke Anbari Jacob Lipke
PDWR according to plan 7.60 7.60 7.60 6.10 5.20 6.00
PDWR will follow current SPI trend 8.75 8.75 8.75 4.38 4.38 5.25
PDWR will follow current SCI trend 9.30 9.11 9.11 4.65 4.46 5.39
Assesment indicator · 1.15 1.15 · 0.55 0.75

100% correlation of all methods to the following straight- for formal project performance reporting. All calculations
forward reasons: and graphs are done by use of a Microsoft Excel spread-
sheet. The different schedule forecasting methods will be
1. All methods apply the same basic parameters such as applied to real project data for three projects. Each project
EV, PD, PV, . . . has a different performance, one project is behind schedule
2. All methods use linear formulas. but under cost, one project is late with a cost over-run and
3. The planned values are linear as well. one project is ahead of schedule but with a cost over-run.
The real-life data of the three projects is summarized in
One could conclude that the three schedule forecasting Table 5.
methodologies have equal validity. However, in a real Project 1. Re-vamp check-in: The project concerns a
project environment it is seldom true that the planned revamping of different check-in islands. This project existed
value is linear (but rather it has the notorious S-shaped mainly out of electrical works (engineering, installation and
curve). Instead, one can assume a learning curve factor commissioning) and automation works (programming,
to denote that work efficiency increases over time due implementing and commissioning). The planned duration
to experience and other beneficial factors. Learning was 9 months, with a budget at completion of € 360.738.
curves have been studied in literature from a project For detailed project data, we refer to Table 6 of the Appen-
scheduling and monitoring point-of-view in [12–18]. The dix A. Fig. 3 displays the different earned value metrics.
right part of Fig. 2 shows the non-linear PV rate and The project was delivered 4 months later than expected,
Table 4 displays the calculated forecasting metrics. As but under budget.
a result, the forecasted durations are no longer identical, The graph of the SV and SV(t) along the project dura-
but depend on the used method. In our example, the tion (the left upper graph of Fig. 3) reveals that the SV fol-
earned schedule method results in the longest forecasted lows a negative trend till February 2003, followed by a
project durations. Jacob and Kane [11] suggest to use of positive trend and finally ending with a zero variation.
smaller time increments for the reporting periods to The SV(t) graph, on the contrary, shows a negative trend
approximate a linear model, reducing the possible result- along the complete project duration, and ends with a
ing errors. cumulative variation of 4 months, which is exactly the
projectÕs delay. A similar effect is revealed in the graph of
3.2. Forecasting at project level the schedule performance metrics (the left middle of
Fig. 3). During the early and middle stages, both SPI and
The illustrations and results of this section are drawn SPI(t) correlate very well. However, towards the late pro-
from a simplified earned value management approach for ject stage (at the ca. 75% completion point), the SPI
managing complex system projects of an airport luggage becomes unreliable showing an improving trend while the
handling systems at Fabricom Airport Systems in Brussels project is slipping further away. This further performance
(Belgium). Weekly meetings with the project team provide decline is clearly shown by the SPI(t) indicator.
the progress data, which is then translated into earned The forecast of the three different schedule forecasting
value metrics, according to the pre-defined earned value methods have been applied and displayed at the right of
methods. The data is then rolled-up to monthly values Fig. 3. The graph reveals some repetitive patterns, regardless

Table 5
Our real-life project data for three project at Fabricom Airport Systems
Project Category Budget at completion Cost at completion Planned duration Actual duration
1 Revamp check in Late finish cost under-run € 360,738 € 349,379 9 13
2 Link lines Late finish cost over-run € 2,875,000 € 3,247,000 9 12
3 Transfer platform Early finish cost over-run € 906,000 € 932,000 10 9
296 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

SV SV(t) P.F. = 1

0 0.0 13.0

-10,000 12.0
Thousands of Euro

-1.0
-20,000

Months
11.0

Months
-30,000 -2.0
10.0
-40,000
-3.0 9.0
-50,000
8.0
-60,000 -4.0 Jun- Jul- Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun-
Jun- Jul- Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun- 02 02 02 02 02 02 02 03 03 03 03 03 03
02 02 02 02 02 02 02 03 03 03 03 03 03
EAC(t)PV1 EAC(t)ED1 EAC(t)ES1

SPI SPI(t) P.F. = SPI, SPI(t)

1.00 13.0
0.95
12.0
0.90

Months
0.85 11.0
0.80
10.0
0.75
0.70 9.0
0.65
8.0
0.60 Jun- Jul- Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun-
Jun- Jul-02 Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun- 02 02 02 02 02 02 02 03 03 03 03 03 03
02 02 02 02 02 02 03 03 03 03 03 03
EAC(t)PV2 EAC(t)ED2 EAC(t)ES2

TCSPI TCSPI(t) TCSPI - 11 P.F. = SCI, SCI(t)

TCSPI(t) - 11 TCSPI - 13 TCSPI(t) - 13
13.0
1.2 12.0
Months

1.0 11.0

0.8 10.0

0.6 9.0

0.4 8.0
Jun- Jul- Aug- Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun-
0.2 02 02 02 02 02 02 02 03 03 03 03 03 03
Jun-02 Jul-02 Aug-02 Sep-02 Oct-02 Nov-02 Dec-02 Jan-03 Feb-03 Mar-03 Apr-03 May-03 Jun-03
EAC(t)PV3 EAC(t)ED3 EAC(t)ES3

Fig. 3. The earned value metrics for Project 1 ‘‘Re-vamp check-in’’ (late finish, under budget).

of the scenario (see Table 1). First, all methods correlate both indices produce similar results. Since there were no
very well during the early and middle project stages, and signs of an improved schedule efficiency at the September
produce nearly similar results. Second, the earned schedule 2002 project review, it was decided to take a two-months
method clearly outperforms all other methods during the project delay into account (revised project dura-
last project stage reporting periods. Finally, the graphs dis- tion = 9 + 2 = 11 months). From this point onwards,
play bizarre and unreliable results for the planned value the new TCSPI indicators (referred to as ‘‘TCSPI-11’’
rate method once the planned time at completion has been and ‘‘TCSPI(t)-11’’) indicators have been computed.
reached, and is therefore not a good duration predictor. After 7 months (the December 2002 project review),
The graphs also reveal that the earned schedule method the TCSPI-11 indicator show a declining trend, indicat-
always forecasts a higher project duration, for each of ing that a lower performance efficiency is needed. How-
the three scenarios. Moreover, both methods are quasi ever, the TCSPI(t)-11 indicator just started an upward
un-sensitive to the scenarios, which might be explained trend, which is a clear indication that improved perfor-
by the fact that the bad schedule performance (late finish) mance is crucial to finish the project within the revised
is compensated by a good cost performance (cost under- deadline of 11 months. A revised scenario to allow for
run). a 4-months delay resulted in a revised targeted project
The graph in the lower right part of Fig. 3 shows the duration of 13 months (with new indicators ‘‘TCSPI-
evolution of the Ôto complete schedule performance 13’’ and ‘‘TCSPI(t)-13’’). The TCSPI-13 continuously
indexÕ (which are defined in the earned schedule and shows a lower value compared to the TCSPI(t)-13.
the earned duration method) over time. These indices Project 2. Link lines: Table 7 of the Appendix A displays
show the performance needed to complete the project the data of the ‘‘link lines’’ project, which links two piers
on time and is given by TCSPI (calculated with the with fully automated baggage conveying lines. The planned
earned duration method) and TCSPI(t) (calculated with duration was 9 months, and the project finished 3 months
the earned schedule method). At the early project stage, later with a cost over-run. Fig. 4 displays the forecasted
 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302 297

SV SV(t) P.F. = 1

0 0.0 13
12
-100 -0.5
12

Months
-200 -1.0 11

Months
Euro

-300 -1.5 11
10
-400 -2.0
10
-500 -2.5 9
-600 -3.0 Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun- Jul-04 Aug-
03 03 03 03 04 04 04 04 04 04 04
Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun- Jul-04 Aug-
03 03 03 03 04 04 04 04 04 04 04
EAC(t)PV1 EAC(t)ED1 EAC(t)ES1

SPI SPI(t) P.F. = SPI, SPI(t)

1.0 14
14
13
0.9 13

Months
12
12
0.8 11
11
10
0.7 10
9
0.6 Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun- Jul-04 Aug-
03 03 03 03 04 04 04 04 04 04 04
Sep- Oct-03 Nov- Dec- Jan-04 Feb-04 Mar-04 Apr-04 May- Jun-04 Jul-04 Aug-
03 03 03 04 04 EAC(t)PV2 EAC(t)ED2 EAC(t)ES2

TCSPI TCSPI - 11 TCSPI - 12 TCSPI(t) P.F. = SCI, SCI(t)

TCSPI(t) - 11 TCSPI(t) - 12 15
1.4
14
1.2 13
Months

1.0 12

11
0.8
10
0.6
9
0.4 Sep- Oct- Nov- Dec- Jan- Feb- Mar- Apr- May- Jun- Jul-04 Aug-
Sep-03 Oct-03 Nov- Dec-03 Jan-04 Feb-04 Mar-04 Apr-04 May- Jun-04 Jul-04 Aug- 03 03 03 03 04 04 04 04 04 04 04
03 04 04
EAC(t)PV3 EAC(t)ED3 EAC(t)ES3

Fig. 4. The earned value metrics for Project 2 ‘‘Link lines’’ (late finish, over budget).

SV SV(t) P.F. = SCI, SCI(t)

11.0
80 2.0
10.5
60 1.5 10.0
Months

9.5
Months
kEuro

40 1.0
9.0

20 0.5 8.5

8.0
0 0.0 Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03
Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03
EAC(t)PV3 EAC(t)ED3 EAC(t)ES3

SPI SPI(t) P.F. = SPI, SPI(t)

11.0
1.20
10.5
1.15 10.0
Months

9.5
1.10
9.0

1.05 8.5

8.0
1.00 Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03
Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03 EAC(t)PV2 EAC(t)ED2 EAC(t)ES2

TCSPI TCSPI(t) P.F. = 1

1.2 11.0

1.0 10.5

0.8
10.0
Months

9.5
0.6
9.0
0.4
8.5
0.2
8.0
0.0 Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03
Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03
EAC(t)PV1 EAC(t)ED1 EAC(t)ES1

Fig. 5. The earned value metrics for Project 3 ‘‘Transfer platform’’ (early finish, over budget).
298 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

results in a similar way as Fig. 3. The graphs reveal that the project monitoring in the early and middle stages. How-
forecasting methods correlate well for the first two thirds of ever, we recommend to shift to the earned schedule
the project, and show a better performance of the earned method for monitoring project progress at the final stage
schedule method towards the end of the project. A similar of the project. Moreover, we recommend to use these
behaviour is reflected in the Ôto complete schedule perfor- schedule forecasting methods at least at the cost account
mance indicesÕ, producing similar results during the early level or at higher levels of the work breakdown structure.
and middle project stages, and an outperforming accuracy This is contradictory to the statements given by Jacob [6]
for the TCSPI(t)-12 index at the end of the project. As a who argues that the schedule forecast metrics should
contrast, the TCSPI-12 shows a decreasing trend (less per- only be used at the level of the activity. Although we
formance is needed) whilst the project is slipping further recognize that, at higher WBS levels, effects (delays) of
away (see Fig. 5). non-performing activities can be neutralized by well per-
Project 3. Transfer platform: The third project is a forming activities (ahead of schedule), which might result
renovation of the transfer baggage conveying system in masking potential problems, we believe that this is the
due to changed baggage flows and security issues. This only approach that can be taken by practitioners.
project had a planned duration of 10 months, while it Indeed, the earned value metrics are set-up as early
finished within 9 months, with a cost over-run (see Table warning signals to detect in an easy and efficient way
8 for details). The graphs confirm the results found by (i.e. at the cost account level, or even higher), rather
the previous projects. At the late project stage, the than a simple replacement of the critical-path based
planned value rate and the earned duration method give scheduling tools. This early warning signal, if analyzed
more pessimistic (i.e. longer duration) results and the properly, defines the need to eventually ‘‘drill-down’’ into
TCSPI metrics produce higher values than the TCSPI(t). lower WBS-levels. In conjunction with the project sche-
The overestimation of duration and/or the needed effi- dule, it allows to take corrective actions on those activi-
ciency calculated by the earned duration method may ties which are in trouble (especially those tasks which are
cause wrong decision-taking by the upper management. on the critical path). Our forecasting results on the three
real-life projects demonstrate that forecasting project
4. Recommendations and conclusions duration with earned value metrics at higher WBS levels
provides reliable early warning signals.
In this paper, we compared three different project dura- Our future research intensions are 3-fold. In order to
tion methods using earned value metrics and evaluate generalize the results found in this study, we will test the
them on fictive and real-life project data. We presented three earned value concepts (planned value, earned sche-
a generic formula to forecast the duration of a project dule, earned duration) on projects based on a full-factorial
and linked them to different project situations. Each simulation experiment, rather than relying on a (small) set
method can be further sub-divided into three different of real-life project. Secondly, we aim at combining different
forecasting models as a function of the project situation. methods depending on the risk profile and other character-
We applied each method on a fictive single-activity project istics of the project. Finally, we want to link the forecasting
with linear and non-linear increasing periodic values methods to their corresponding corrective actions that can
reflecting the absence or presence of learning curves as be taken. To that purpose, we will rely and extend the work
well as three real-life project from Fabricom Airport Sys- done by Lipke [19].
tems, Belgium. We summarized the often confusing termi- As a final remark, we cite the letter to the editor of Har-
nology of the different methods in two tables. vard Business Review from Cooper [20] as a response to
The results show a similar forecasting accuracy for each the article written by Fleming and Koppelman [3]. In this
method in the linear planned value case. However, the letter, the author argues that the use of earned value man-
introduction of learning curves, which is much more realis- agement can be questioned when they are applied in highly
tic in the project world, results in a different forecasting complex projects. Due to the cycles of rework, the accuracy
accuracy for the three methods. The three real-life projects of the EVM metrics can be biased, leading to incorrect
reveal that the earned schedule method was the only management decisions. It is our ultimate goal to investigate
method which showed satisfying and reliable results during this research topic and provide an answer on this issue. In
the whole project duration. Consequently, the results con- doing so, we will rely and extend the partial answer formu-
firm the previously found results that the results obtained lated by Lipke [10] who measures the Ôeffective earned
by the planned value rate and the earned duration method valueÕ when the project is the subject of a vast amount of
are unreliable at the end of the project. Instead, the earned rework cycles.
schedule method seems to provide valid and reliable results
along the projectÕs lifespan.
As a conclusion, we believe that the use the planned Appendix A
value method, the earned duration method or the earned
schedule method depending on the need and knowledge Detailed information about the three real-life project at
of the project manager might lead to similar results for Fabricom Airport systems (see Tables 6–8).
Table 6
Detailed information for project 1 ‘‘Re-vamp check-in’’
Jun-02 Jul-02 Aug-02 Sep-02 Oct-02 Nov-02 Dec-02 Jan-03 Feb-03 Mar-03 Apr-03 May-03 Jun-03
AC 25,567 66,293 78,293 124,073 191,367 259,845 285,612 290,843 303,489 316,431 320,690 336,756 349,379

S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

EV 25,645 68,074 89,135 125,244 198,754 268,763 292,469 306,725 312,864 327,694 338,672 349,861 360,738
PV 28,975 81,681 91,681 138,586 218,141 302,478 323,632 345,876 360,738 360,738 360,738 360,738 360,738

SV 3330 13,607 2546 13,342 19,387 33,715 31,163 39,151 47,874 33,044 22,066 10,877 0
CV 78 1781 10,842 1171 7387 8918 6857 15,882 9375 11,263 17,982 13,105 11,359
SPI 0.885 0.833 0.972 0.904 0.911 0.889 0.904 0.887 0.867 0.908 0.939 0.970 1.000
CPI 1.003 1.027 1.138 1.009 1.039 1.034 1.024 1.055 1.031 1.036 1.056 1.039 1.033
SCI 0.888 0.856 1.107 0.912 0.946 0.919 0.925 0.935 0.894 0.941 0.991 1.008 1.033

AD 1 2 3 4 5 6 7 8 9 10 11 12 13
ES 0.885 1.742 2.745 3.716 4.756 5.600 5.881 6.201 6.491 7.183 7.676 8.268 9.000
PD 9 9 9 9 9 9 9 9 9 9 9 9 9

SV(t) 0.115 0.258 0.255 0.284 0.244 0.400 1.119 1.799 2.509 2.817 3.324 3.732 4.000
SPI(t) 0.885 0.871 0.915 0.929 0.951 0.933 0.840 0.775 0.721 0.718 0.698 0.689 0.692
SCI(t) 0.888 0.894 1.042 0.938 0.988 0.965 0.860 0.817 0.743 0.744 0.737 0.716 0.715

EAC(t)PV1 9.083 9.339 9.064 9.333 9.484 9.841 9.777 9.977 10.194 9.824 9.551 9.271 9.000
EAC(t)ED1 9.115 9.333 9.083 9.385 9.444 9.669 9.674 9.906 10.194 10.916 11.673 12.362 13.000
EAC(t)ES1 9.115 9.258 9.255 9.284 9.244 9.400 10.119 10.799 11.509 11.817 12.324 12.732 13.000

EAC(t)PV2 10.169 10.799 9.257 9.959 9.878 10.129 9.959 10.149 10.377 9.908 9.586 9.280 9.000
EAC(t)ED2 10.169 10.799 9.257 9.959 9.878 10.129 9.959 10.149 10.377 11.008 11.717 12.373 13.000
EAC(t)ES2 10.169 10.334 9.835 9.689 9.461 9.642 10.712 11.611 12.479 12.530 12.897 13.062 13.000

EAC(t)PV3 10.14 10.52 8.13 9.87 9.51 9.79 9.73 9.62 10.07 9.57 9.08 8.93 8.72
EAC(t)ED3 10.14 10.57 8.50 9.90 9.70 9.99 9.89 10.04 10.34 10.97 11.68 12.36 13.00
EAC(t)ES3 10.14 10.12 9.00 9.64 9.30 9.52 10.62 11.42 12.37 12.44 12.80 13.02 13.00

TCSPI 1.014 1.048 1.014 1.077 1.111 1.223

TCSPI(t) 1.014 1.037 1.042 1.057 1.061 1.133

TCSPI-11 0.741 0.734 0.669 0.635 0.597

TCSPI(t)-11 0.707 0.680 0.780 0.933 1.255

TCSPI-13 0.381 0.299 0.305 0.336 0.362

TCSPI(t)-13 0.560 0.627 0.606 0.662 0.732

299
 300
Table 7
Detailed information for project 2 ‘‘Link Lines’’ (costs in thousands of €)
Sep-03 Oct-03 Nov-03 Dec-03 Jan-04 Feb-04 Mar-04 Apr-04 May-04 Jun- 04 Jul-04 Aug-04

S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

AC 344 452 796 1056 1562 1922 2256 2451 2676 2925 3138 3247
EV 325 427 735 1025 1453 1774 2024 2190 2356 2565 2735 2875
PV 375 525 850 1355 1768 2125 2452 2625 2875 2875 2875 2875
SV 50 98 115 330 315 351 428 435 519 310 140 0
CV 19 25 61 31 109 148 232 261 320 360 403 372
SPI 0.867 0.813 0.865 0.756 0.822 0.835 0.825 0.834 0.819 0.892 0.951 1.000
CPI 0.945 0.945 0.923 0.971 0.930 0.923 0.897 0.894 0.880 0.877 0.872 0.885
SCI 0.819 0.768 0.798 0.734 0.764 0.771 0.741 0.745 0.721 0.782 0.829 0.885

AD 1 2 3 4 5 6 7 8 9 10 11 12
ES 0.867 1.347 2.646 3.347 4.237 5.017 5.717 6.199 6.706 7.653 8.440 9.000
PD 9 9 9 9 9 9 9 9 9 9 9 9
SV(t) 0.133 0.653 0.354 0.653 0.763 0.983 1.283 1.801 2.294 2.347 2.560 3.000
SPI(t) 0.867 0.674 0.882 0.837 0.847 0.836 0.817 0.775 0.745 0.765 0.767 0.750
SCI(t) 0.819 0.636 0.814 0.812 0.788 0.772 0.733 0.692 0.656 0.671 0.669 0.664

EAC(t)PV1 9.157 9.307 9.360 10.033 9.986 10.099 10.340 10.362 10.625 9.970 9.438 9.000
EAC(t)ED1 9.133 9.373 9.406 9.974 9.891 9.991 10.222 10.326 10.625 11.078 11.536 12.000
EAC(t)ED1 9.133 9.653 9.354 9.653 9.763 9.983 10.283 10.801 11.294 11.347 11.560 12.000
EAC(t)PV2 10.385 11.066 10.408 11.898 10.951 10.781 10.903 10.788 10.983 10.088 9.461 9.000
EAC(t)ED2 10.385 11.066 10.408 11.898 10.951 10.781 10.903 10.788 10.983 11.209 11.563 12.000
EAC(t)ES2 10.381 13.363 10.204 10.756 10.621 10.763 11.020 11.615 12.079 11.760 11.730 12.000
EAC(t)PV3 10.99 11.71 11.27 12.26 11.77 11.68 12.15 12.07 12.47 11.50 10.35 10.16
EAC(t)ED3 10.93 11.60 11.02 12.14 11.40 11.18 11.35 11.12 11.25 11.38 11.65 12.00
EAC(t)ES3 10.93 14.03 10.80 10.96 11.04 11.16 11.48 12.05 12.50 12.01 11.84 12.00

TCSPI 1.017 1.053 1.068 1.195 1.223 1.330

TCSPI(t) 1.017 1.093 1.059 1.131 1.191 1.328
TCSPI-11 0.815 0.798 0.805 0.775 0.812
TCSPI(t)-11 0.794 0.797 0.821 0.934 1.147
TCSPI-12 0.581 0.542 0.539 0.536
TCSPI(t)-12 0.700 0.765 0.674 0.560
Table 8

S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

Detailed information for project 3 ‘‘Transfer line’’ (costs in thousands of €)
Mar-03 Apr-03 May-03 Jun-03 Jul-03 Aug-03 Sep-03 Oct-03 Nov-03 Dec-03
AC 35 95 174 412 623 754 874 932 952 952
EV 36 93 169 402 597 735 839 887 906 906
PV 34 87 157 373 549 673 798 842 876 906
SV 2 6 12 29 48 62 41 45 30 0
CV 1 2 5 10 26 19 35 45 46 46
SPI 1.059 1.069 1.076 1.078 1.087 1.092 1.051 1.053 1.034 1.000
CPI 1.029 0.979 0.971 0.976 0.958 0.975 0.960 0.952 0.952
SCI 1.089 1.046 1.046 1.052 1.042 1.065 1.009 1.003 0.984

AD 1 2 3 4 5 6 7 8 9 10
ES 1.038 2.086 3.056 4.165 5.387 6.496 7.912 9.367 10.000
PD 10 10 10 10 10 10 10 10 10
SV(t) 0.667 0.086 0.056 0.165 0.387 0.496 0.912 1.367 1.000
SPI(t) 1.038 1.043 1.019 1.041 1.077 1.083 1.130 1.171 1.111
SCI(t) 1.067 1.021 0.989 1.016 1.032 1.055 1.085 1.114 1.057

EAC(t)PV1 9.994 9.981 9.962 9.909 9.850 9.806 9.872 9.859 9.906
EAC(t)ED1 9.941 9.862 9.771 9.689 9.563 9.447 9.640 9.572 9.692
EAC(t)ES1 9.962 9.914 9.944 9.835 9.613 9.504 9.088 8.633 9.000
EAC(t)PV2 9.444 9.355 9.290 9.279 9.196 9.156 9.511 9.493 9.669
EAC(t)ED2 9.444 9.355 9.290 9.279 9.196 9.156 9.511 9.493 9.669
EAC(t)ES2 9.636 9.589 9.818 9.604 9.281 9.236 8.848 8.541 9.000
EAC(t)PV3 9.182 9.556 9.565 9.509 9.596 9.393 9.908 9.974 10.160
EAC(t)ED3 9.210 9.513 9.476 9.410 9.379 9.238 9.616 9.568 9.703
EAC(t)ES3 9.396 9.752 10.020 9.744 9.468 9.320 8.925 8.568 9.000

TCSPI 0.993 0.983 0.967 0.948 0.913 0.862 0.880 0.786 0.692
TCSPI(t) 0.996 0.989 0.992 0.973 0.923 0.876 0.696 0.317 0.000

301
302 S. Vandevoorde, M. Vanhoucke / International Journal of Project Management 24 (2006) 289–302

References [11] Jacob DS, Kane M. Forecasting schedule completion using earned
value metrics revisited. The Measurable News 2004(Summer):1. 11–
7.
[1] Anbari F. Earned value method and extensions. Project Manage J
[12] Amor JP. Scheduling programs with repetitive projects using com-
2003;34(4):12–23.
posite learning curve approximations. Project Manage J 2002;33(2):
[2] Fleming Q, Koppelman J. Earned value project management New-
16–29.
towns Square. PA: PMI; 2000.
[13] Amor JP, Teplitz CJ. Improving CPMÕs accuracy using learning
[3] Fleming Q, Koppelman J. WhatÕs your projectÕs real price tag?
curves. Project Manage J 1993;24(4):15–9.
Harvard Bus Rev 2003;81(9):20–1.
[14] Amor JP, Teplitz CJ. An efficient approximation procedure for project
[4] Lipke W. Schedule is different. The Measurable News 2003(March):
composite learning curves. Project Manage J 1998;29(3):28–42.
31–4.
[15] Badiru AB. Incorporating learning curve effects into critical resource
[5] Christensen DS. The estimate at completion problem: a review of
diagramming. Project Manage J 1995;2(2):38–45.
three studies. Project Manage J 1993;24:37–42.
[16] Lam KC, Lee D, Hu T. Understanding the effect of the learning–
[6] Jacob D. Forecasting project schedule completion with earned value
forgetting phenomenon to duration of projects construction. Int J
metrics. The Measurable News 2003(March):1. 7–9.
Project Manage 2001;19:411–20.
[7] Henderson K. Earned schedule: a breakthrough extension to earned
[17] Shtub A. Scheduling of programs with repetitive projects. Project
value theory? A retrospective analysis of real project data. The
Manage J 1991;22(4):49–53.
Measurable News 2003(Summer):13–7. 21.
[18] Shtub A, LeBlanc LJ, Cai Z. Scheduling programs with repetitive
[8] Henderson K. Further developments in earned schedule. The Mea-
projects: a comparison of a simulated annealing, a genetic and a pair-
surable News 2004(Spring):15–6. 20–2.
wise swap algorithm. Eur J Oper Res 1996;88:124–38.
[9] Henderson K. Earned schedule in action. The Measurable News
[19] Lipke W. Deciding to act. Crosstalk 2003(December):22–4.
2005:23–8. 30.
[20] Cooper KG. Your projects real price tag – letters to the editor.
[10] Lipke W. Connecting earned value to the schedule. The Measurable
Harvard Bus Rev 2003;81(12):122.
News 2004(Winter):1. 6–16.