Project Control –

Earned Value Management and

Earned Schedule
Infrastructure Management
Dr. Mark Hsu

1
 MTR Case
❑ What now for Hong Kong’s over-budget Sha-Tin-
Central Link (Over HK$99.1 Billion)? And why
the MTR’s numbers did not tell the full story
❑ http://www.scmp.com/news/hong-
kong/economy/article/2123190/what-now-hong-
kongs-over-budget-sha-tin-central-link-and-why
❑ https://www.scmp.com/news/hong-
kong/transport/article/3046511/partial-opening-
scandal-hit-sha-tin-central-mtr-link-set

2
Today’s class

3
Today’s class

4
 Project control
Objectives: Topics:
▪ Understand the basics Project control using EVM
of project control How to measure progress
using Earned Value Planned vs. actual progress
▪ Using EV for cost and Earned Value Management
schedule control
Forecast cost and completion
▪ Using EV for
Monitoring performance
forecasting

5
 Project control
Problems during project execution
❑ Many project activities consume more than
their estimated time and cost.
❑ Productivity among project teams varies and
affects overall coordination in a project
❑ Changes are introduced to the scope of work
that cause disruption and cost overruns
❑ Dealing with suppliers and subcontractors may
not be easy
❑ Tracking technical progress may be hidden
from the controls

6
 Project control
To execute a project as close to as-planned as
possible, we need to:
❑ Accurately follow the project plan
❑ Update the project plan based on new
circumstances
❑ Monitor execution and keep track of resources
❑ Provide progress reports, comparing actual vs.
planned
❑ Forecast cost at completion at any project
stage
❑ Take corrective actions at any stage to bring
time and cost closer to plan

7
 Project control

PROJECT CONTROL

Measuring Cost and Forecasting Schedule

Work Progress Schedule Control at Completion Updating

8
 What is EVM?
❑ A systematic approach to the integration and
measurement of cost, schedule, and technical
(scope) accomplishments on a project.
❑ EVM is a project management tool that integrates
schedule and cost parameters of the contract
❑ All work is planned, scheduled, and budgeted in
time-phased “planned value” increments.
❑ As work is performed it is controlled against the
baseline.
❑ Provides the ability to examine detailed schedule
information, critical program and technical
milestones, and cost data.

9
 Genesis and evolution of EVM
❑ 1960s – DoD adopted Cost/Schedule Control Systems Criteria
(C/SCSC) as an objective measure of progress.
❑ 1970s – Continued use in DoD as a means to offset
cost/schedule risk in cost-plus contracts.
High-tech, newly-developed weaponry
Arms race induced critical schedule needs
❑ 1990s – Policy moved Earned Value into all Federal agencies
OMB Circular A-11
NASA Policy Directive 9501.3
DOE order 413.3
❑ 2003 – OMB began enforcement in all civilian agencies.
(Note: OMB refers to Office of Management and Budget.)

10
 Why do we care?
❑ Projects over budget and behind schedule*
53% of IT projects finish over budget and behind
schedule
52% finish at 189% of their initial budget
18% are simply never completed
❑ Consistent performance measures allow comparison
across the portfolio
❑ Statistical performance projections

*Source: Standish Group’s Chaos Chronicles 2004

11
 Why do we care?
Fundamental difference with traditional
management is the data used for analysis.
Budget vs. Actuals
Earned Value Analysis

12
 Earned Value Management
EVM is:
Performance measurement
• How am I doing against my baseline plan?
Performance management
• What do I need to do to bring the project in on
cost and schedule?

13
 Where can EVM be applied?
EVM suitable for project that have:
Clear definition of work scope
Project schedule range from a few months to
many years
Small to very large cost
EVM not suitable for:
Projects without a clear work definition or
deliverable products

14
 Rudiments of EVM
Based on something called Earned Value:
The concept that task activities earn value as
work progress
A data point that expresses the value of work
accomplished
Earned value (EV) = Quantity done * Unit cost

15
 Rudiments of EVM
Absolutely dependent on the Work Breakdown Structure (WBS)
New Weather Radar

One WBS per program Antenna and

Mechanical
Systems
Transmitter and
Receiver Systems
Data Processing
Systems

Deliverable-oriented
Computer Application Telecommunications

Organizational responsibility identified Hardware Software Systems

in WBS
Work not in the WBS is out-of-scope
All lower-level elements roll up to the WBS total
Full (and accurate) definition is key
Defined deliverable(s)
Timeframe for delivery of product
Total cost (direct and indirect) to deliver product
Assigned organizational responsibility

16
 How to measure progress
What do we measure progress against?
Performance measurement baseline
Budget that is spread over…
Time to accomplish the scope of …
Work. And against which progress can be measured

“Earned Value”
How much progress did we make against the original
plan
Expressed in money (or sometimes hours)

17
 How to measure progress
Calculating Activities’ % Complete
Units completed: repeated production of easily measured
units of work (i.e. pile driving)
% Completed = units completed / total units
Incremental milestones: applies to activities that include
sequenced subtasks. i.e: Installation of major equipment
Received & Inspect 15%
Setting complete 35%
Alignment complete 50%
Internals installed 75%
Testing completed 90%
Accepted by owner 100%

18
 How to measure progress
Start/Finish: for activities that lack of intermediate
milestones
Supervisor opinion
Cost Ratio: for activities that are budgeted based on a
bulk allocation of $ and involve long time and
continuity (i.e: project management, QC & QA, etc)
%Complete = Actual cost (or hours) of work to
date / Forecast @ completion

19
 How to measure progress
Weighed or Equivalent Units:
Applies to activities that are composed of two or
more overlapping subtasks, each with a different
unit of work measurement.
e.g. structural steel with a total weight of 520 tons.
All subtasks are converted into equivalent steel
tons as a unified measurement unit.

20
How to measure progress

21
 Planned vs. actual progress
100%
Direct + indirect
costs Early Start

S-Curve Method Base or

average
Late Start

Actual

Time

❑ Drawback is that we do not know if extra expenses are caused by

fast execution of more work items than planned or simply because
of higher rates paid to execute less work than planned: No indication
on the amount of work for which these expenses were incurred.

22
 Planned vs. actual progress
Time-Based S-Curve Method
Planned Cost
100%
Actual Cost
Cost (Direct + indirect)

Work
hours
Planned 100%
Work

Actual Hours

Time

Now

23
 Planned vs. actual progress
Under Budget / Behind Schedule

24
 Life without EVM
Given:
Total budget of $100,000
12 month effort
Produce 20 units
Status:
Spend to date: $64,000
Time elapsed: 6 months
Units produced: 8 complete, 2 partial
How are you doing, and how do you know how you are
doing? How far along are you?
64% spent (cost)
50% spent (time)

25
 Life with EVM
Real Status:
You should have completed 50% of the work
• You’ve completed 42% of the work
A major vendor was late in delivering certain
parts
• You’ve spent 64% of your budget
You’ve isolated the main cost variance to
aluminum prices
• You’re forecasting additional cost overruns for the
remaining work
• You now know what the drivers are
• You can now take the appropriate actions

26
EVM gives insight

27
 Planned Value
Data source is the Integrated Project Baseline
How much did you expect to have done at point X ?
• Expressed in dollars (PV)
How much do you expect to have done at completion ?
• Budget at Completion (BAC)
$2,500,000

$2,000,000 BAC $2,000,000

$1,500,000
Planned Value
$1,000,000

$500,000

$0
r

r
l

l
ct

ct
n

n
Ap

Ju

Ap

Ju
Ja

Ja
O

28
 Actual Cost
Data source is the Earned Value Management System
The dollar amount actually spent to date.
Has no relationship to work accomplished

$2,500,000
AC = $1,041,000
$2,000,000

$1,500,000
Planned Value
Actual Cost
$1,000,000

$500,000
PV = $1,000,000
$0
t

t
r

r
n

n
l

l
Oc

Oc
Ap

Ap
Ju

Ju
Ja

Ja

29
 Earned Value
Data source is the Earned Value Management System
The dollar value of actual accomplishments to date.
Does not address the money spent in getting there.

$2,500,000

$2,000,000 $2,000,000
PV = $1,000,000
$1,500,000
Planned Value
Earned Value
$1,000,000

$500,000
EV = $851,000

$0
May

May
Nov

Nov
Mar

Mar
Jan

Sep

Jan

Sep
Jul

Jul

30
 Earned Value
Several ways to “earn” it:
➢Percent Completion Estimates

➢Start/Finish of WBS Elements

25% 75% 50% 50% 0% 100%

➢Milestones Completed (0-100%)

➢Percent Completion Estimate with Milestone Gates

31
33% 67% 100%
31
Integrated Performance Report
2,500,000

2,000,000
AC = $1,040,979
1,500,000 Planned Value
Actual Cost
1,000,000 Earned Value
PV = $1,000,000

500,000
EV = $851,000
0
May

May
Nov

Nov
Mar

Mar
Jan

Sep

Jan

Sep
Jul

Jul

32
 Earned Value Management
Basic Elements

33
 Earned Value Management
$
BCWS
(PV)

ACWP
(AC)
Cost
Variance
Schedule
Variance

BCWP (EV)

Now Time

Time Variance u EV = Earned Value

u AC = Actual Cost
u PV = Planned Value

34
 Monitoring performance
Common Measures
Cost Variance
CV = EV – AC = BCWP - ACWP
Schedule Variance
SV = EV – PV = BCWP - BCWS
Cost Performance Index
CPI = EV / AC = BCWP / ACWP
Schedule Performance Index
SPI = EV / PV = BCWP / BCWS
Critical Ratio
CR = SPI x CPI

35
Schedule variance

36
Cost variance

37
Variance at completion

38
Monitoring variances

39
 Cost performance
Cost Variance (EV-AC)
2,500,000 $851K - $1041K = -$190K
AC = $1,040,979 CPI (EV/AC) = 0.8175
PV = $1,000,000
2,000,000
EV = $851,000

1,500,000 Planned Value

Actual Cost
1,000,000 Earned Value

500,000

0
May

May
Nov

Nov
Mar

Mar
Jan

Jan
Jul

Jul
Sep

Sep

40
 Schedule performance
Schedule Variance (EV-PV)
$851K - $1,000K = - $149K
2,500,000
PV = $1,000,000 SPI (EV/PV) = 0.851
AC = $1,040,979
2,000,000
EV = $851,000

1,500,000 Planned Value

Actual Cost
1,000,000 Earned Value

500,000 And look at this:

$ - 149K = 1 Month Slippage
0
Nov

Nov
Jan

Jan
Jul

Jul
Mar

Mar
Sep

Sep
May

May

41
 Monitoring performance
CPI, SPI are basic
If CPI > 1.0
Overestimated amount of effort required
Measurement of % complete optimistic
Better than planned productivity
If CPI <1.0
Underestimated amount of work required
Measurement of % complete pessimistic
Poorer than planned productivity

42
 Monitoring performance
If SPI > 1.0
Started work out of sequence
Better productivity than planned
If SPI <1.0
Possible understaffing
Delayed starts
Work more difficult than planned
Improper sequencing of activities
Poor work organization / execution

43
 Monitoring performance

EVM allows isolation of

schedule and cost variance

44
Monitoring performance

45
 EVM example 1
Suppose that work on a project task was expected to
cost $1500 to complete the task and the workers were
originally scheduled to have finished today. As of today,
however, the workers have actually expended $1350,
and our best estimate is that they are about 2/3 finished.

Cost/Spending Variance = EV – AC =
1500(2/3) – 1350 = -350
Schedule Variance = EV – PV =
1500(2/3) – 1500 = -500
CPI = EV/AC = 1500(2/3)/1350 = 0.74
SPI = EV/PV = 1500(2/3)/1500 = 0.67
46
Forecasting

47
 Forecasting
Maintaining same cost
performance index:
EAC
Estimated (Remaining Cost) to
Cost
Variance at Completion
Present
ETC = (BAC - BCWP)/CPI
BCWS Estimated at Completion (Total
Cost)
ACWP EAC = ETC + ACWP
Cost
Variance at
Present
Maintaining same cost variance:
BCWP (EV) Estimated at Completion
Now
EAC = BCWS at completion +
Time
(ACWP – BCWP) at present

48
 Forecasting
Estimate to Complete (ETC)
How much will we have to spend (after today) to
complete?
Total budget less what we’ve earned to date. (With
some assumptions).
• Performance to date has been anomalous
• Cost Performance to date will continue
• Both Cost and Schedule Performance to date
will continue

49
 Forecasting
Estimate to Complete (ETC)
Performance to date has been anomalous:
ETC = BAC – EV = $1,149,000 → BAC = $2,000,000
Cost Performance to date will continue:
ETC = (BAC-EV)/CPI = $1,405,504
Both Cost and Schedule Performances to date will
continue:
ETC = (BAC-EV)/(CPI*SPI) = $1,651,574

PV = BCWS = $1,000,000
CPI = BCWP/ACWP = 0.8175
AC = ACWP = $1,041,000
SPI = BCWP/BCWS = 0.851
EV = BCWP = $851,000

50
 Forecasting
Estimate at Completion (EAC)
❖How much will we have spent when we’re actually
done?
❖Again, based on the same assumptions as ETC.
EAC = AC + (BAC-EV) = $2,190,000
EAC = AC + (BAC – EV) / CPI = $2,446,505
EAC = AC + (BAC – EV) / (CPI*SPI) = $2,691,856

CPI = BCWP/ACWP = 0.8175

SPI = BCWP/BCWS = 0.851

51
 EVM example 1 (continued)
❑ We can use these calculations to determine some
additional items of interest as well, such as the estimated
(remaining cost) to completion (ETC) and the projected
(total cost) estimated at completion (EAC) if nothing is
done to correct the problem. Given that the budget at
completion (BAC) is 1,500 and the EV is 1,500(2∙3) =
1,000
ETC = (BAC − EV)/CPI = (1,500 −1,000)/0.74 = 676
❑ Note that this assumes that the work will be completed at
the same level of efficiency or inefficiency as conducted
thus far. The total cost to complete the task is
EAC=ETC+AC=676+1350=2,026

52
 Forecasting
To-Complete Performance Index (TCPI)
❖How well do we have to perform to get back on
track?
❖Again, based on the same assumptions as ETC.
TCPIC = 1/CPI = 1.22
TCPIS = 1/SPI = 1.17
TCPICS = 1/(CPI*SPI) = 1.437

CPI = BCWP/ACWP = 0.8175

SPI = BCWP/BCWS = 0.851

53
 Prede- Estimated Costs
Activity Description Estimated time
cessor BAC
(weeks)
Determine equipment - 3 1000
A needs
B Obtain vendor proposals - 5 1000
C Select vendor A,B 2 1500
D Order system C 6 2000
Design new warehouse C 5 5000
E layout
F Design warehouse E 3 10000

G Design computer interface C 4 8000

H Computer interface G 3 10000

implementation
I Install system D ,F 4 10000

J Train system operators I,H 2 10000

K Test system J 2 10000

54
Go to Chart Go to Summary
 Example: Monitoring performance
D I

A
start C E F

B
G H
J K

End

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
WEEKS

A
B
C
D
E
F
G
H
I
J
K

55
 Example: Monitoring performance
Field Report, data capture in the 8th Week
Activity Actual Cost ($) % complete # of weeks Remaining
at at out of total Duration
8th Week 8th week
A 1,000 100 3 out of 3 0
B 1,200 100 5 out of 5 0
C 2,000 100 2 out of 2 0
D 1,000 15 1 out of 6 5 weeks
E 1,500 25 1 out of 5 4 weeks
F 0 0 Not started 0
G 0 0 1 out of 4 3 weeks
H 0 0 Not started 0
I 0 0 Not started 0
J 0 0 Not started 0
K 0 0 Not started 0
6,700 56
 1 2 3 4 5 6 7 8

333.33 333.33 Planned

A 3 333.33 3 ($)
100% Actual (%)
B 200 200 200 200 200
100%
C 750 750
100%
D 333.3
15%
E 1000
25%
F
Project Schedule at the 8th week
G (late) 2000
0%

Tot Cost 533.33 533.3 533.33 200 200 750 750 3333
Cum. Cost 533.33 1067 1600 1800 2000 2750 3500 6833 Go to BC

57
 Example: Monitoring performance

(1) (2) (3) (4)=(1)*(3) (5) (6)=(4)-(5) (7)=(4)/(5) (8)=(4)-(2) (9)=(4)/(2) (10)=[(1)-(4)]/(7) (11)=(5)+(10)
Activit
y BAC BCWS PC% BCWP ACWP CV CPI SV SPI ETC EAC
A 1000 1000 100% 1000 1000 0 1 0 1 0 1000

B 1000 1000 100% 1000 1200 -200 0.833 0 1 0 1200

C 1500 1500 100% 1500 2000 -500 0.75 0 1 0 2000
D 2000 333.3 15% 300 1000 -700 0.3 -33.3 0.9 5666.67 6666.7
E 5000 1000 25% 1250 1500 -250 0.833 250 1.25 4500 6000
G 8000 2000 0% 0 0 0 -2000 0

Total 6833 5050 6700 -1650 0.754 -1783 0.739

Over budget - Behind Schedule

Go to BC Go to AC

58
 Problems with EVM
❑ Inadequate early warning
❑ Poor implementation of EVM
❑ Reliable, valid information
❑ Ignores technological progress
❑ Ignores quality

59
 What wrong with EVM?
❑Ignores critical path

BCWS = 7000
BCWP = 7000
SV = 0

60
 What wrong with EVM?
Example with total float ..

SPI = 1.5 ahead of schedule

CPI = 0.75 costing too much
TF = +60 days way ahead on critical activities

More work performed than scheduled and

critical path activities have been expedited. But
… deteriorating productivity.

61
 What wrong with EVM?
Traditional schedule EVM metrics are good at beginning of project
– Show schedule performance trends
But the metrics don’t reflect real schedule performance at the end
– Eventually, all “budget” will be earned as the work is
completed, no matter how late you finish
SPI improves and ends up at 1.00 at end of project
SV improves and ends up at $0 variance at end of project
– Traditional schedule metrics lose their predictive ability over
the last third of project
Impacts schedule predictions, EAC predictions
– Project managers don’t understand schedule performance in
budget descriptions

62
What wrong with EVM?

63
 What wrong with EVM?
Why does this happen?
❑SV = EV – PV = BCWP– BCWS
❑SPI = EV / PV = BCWP / BCWS
At planned completion PV = BCWS = BAC
At actual completion EV = BCWP = BAC
When actual > planned completion
❑SV = BAC– BAC = $0
❑SPI = BAC / BAC = 1.00 Regardless of lateness

64
 Earned schedule concept
❑ Earned Schedule (ES) analysis is a breakthrough
analytical technique that derives schedule
performance measures in units of time, rather than
cost.
❑ The same basic Earned Value Management (EVM)
data points are used. Indicators, similar to those for
cost, are derivable from the earned schedule
measure.

65
 Earned schedule concept
❑ As described by Lipke in the seminal paper (Lipke 2003,
p12):

“The idea of Earned Schedule is analogous to Earned

Value. However, instead of using cost for measuring
schedule performance, we would use time. Earned
Schedule is determined by comparing the cumulative
BCWP earned to the performance baseline, BCWS.
The time associated with BCWP, i.e. Earned Schedule,
is found from the BCWS S-curve.”

*An alternative explanation is that the ES idea is to identify

the time at which the amount of EV accrued should
have been earned.

66
 Earned schedule
❑Required measures
▪ Performance Management Baseline (PMB) – the
time phased planned values (BCWS) from project
start to completion
▪ Earned Value (BCWP) – the planned value which
has been “earned”
▪ Actual Time (AT) - the actual time duration from
the project beginning to the time at which project
status is assessed
❑All measures available from EVM

67
 Earned schedule
Expressed algebraically, ES cum is the number of completed PV
time increments EV exceeds PV plus the fraction of the
incomplete PV increment in the unit of time (i.e. weekly or
monthly) being utilized.
Therefore ES cum = C + I where:
C = number of time increments where EV exceeds PV; and
I = (EV – PVC) / (PVC+1 – PVC)
Since ES is calculated cumulatively, periodic ES is calculated by
simply subtracting the current period ES cumulative (cum) by the
preceding period:
ES period(n) = EScum(n) – EScum(n-1)

68
 Earned schedule
❑ The Earned Schedule Indicators
▪ Schedule Variance (time): SV(t) = ES – AT, where AT
= actual time
▪ Schedule Performance Index (time): SPI(t) = ES / AT

❑ Key Points:
▪ ES Indicators constructed to behave in an
analogous manner to the EVM Cost Indicators, CV
and CPI
▪ SV(t) and SPI(t) not constrained by BCWS
calculation reference
▪ SV(t) and SPI(t) provide duration based measures of
schedule performance

69
 Earned schedule
❑ The Schedule Variance, SV(t), is positive when the
ES exceeds AT, and, of course, is negative when it
lags. The Schedule Performance Index, SPI(t), is
greater than 1.0 when ES exceeds AT, and, of course,
is less than 1.0, when ES is less than AT.
❑ These proposed indicators are completely analogous
to the EVM cost indicators, CV and CPI. The
proposed schedule indicators are referenced to
“actuals,” just as are the EVM cost indicators.

70
 Earned schedule
What happens to the ES indicators, SV(t) &
SPI(t), when the planned project duration (PD) is
exceeded (BCWS = BAC)?

They Still Work …

SV(t) will be negative (time behind schedule)
SPI(t) will be < 1.00

71
Earned schedule

72
Earned schedule example

73
 Earned schedule example
Late finish project (Table 2) …

BCWP August = $1900

BCWS July = $1805
BCWS August = $2135
ES = 7 + (1900-1805)/(2135-1805) = 7.288 months
SV(t) = ES – AT = 7.288 - 8 = -0.722 months
SPI(t) = 7.288 / 8 = 0.911

74
 Earned schedule example
BCWP Dec = 2555
BCWS Sep = 2435
(BCWS Oct, Nov, and Dec > BCWP Dec)
BCWS Oct = 2665
ES = 9 + (2555-2435)/(2665-2435) = 9.522 months
SV(t) = ES – AT = 9.522 – 12 = -2.478 months
SPI(t) = 9.522/12 = 0.794

75
Schedule variance comparison

76
Schedule performance index comparison

77
Earned Schedule VS Earned Value Indicators

78
79
 ES Applied to Real Project Data
❑Late Finish Project Analysis
No EVM data prior to week 11
SV($) and SV(t) show strong correlation until week 19
• Week 20 (The week of the project’s scheduled
completion) Client delay halted project progress until
resolution in Week 26
SV($) static at -$17,500 in spite of schedule delay

❑Before trending to $0 at project completion

SV(t) correctly calculates and displays
• Week on week schedule delay
• Project -14 week schedule delay at completion

80
81
 Early finish project analysis
❑ This project completed 3 weeks ahead of schedule
▪ In spite of externally imposed delay between
weeks 16 and 19
❑ SV($) and SV(t) show strong correlation over life of
project
▪ Including the delay period
▪ SV(t)’s advantage is calculating delay as a
measure of duration
❑ With Early Finish projects
▪ ES metrics SV(t) and SPI(t) have behaved
consistently with their historic EVM counterparts

82
ES – schedule analysis

83