2012 AACE INTERNATIONAL TRANSACTIONS

CDR.980

Proving Engineering Productivity Loss

Dr. Tong Zhao, PE PSP and J. Mark Dungan
ABSTRACT—The measured mile/baseline method has been widely accepted to quantify labor
productivity loss, but it has not been successfully used in engineering productivity because of the
challenges in engineering productivity measurement and the determ ination of productivity
benchmark. This paper presents a series o f procedures based on the measured mile/baseline
method to quantify engineering productivity loss from a project specific perspective. A case study
on the piping discipline in a large scale process plant project is used to illustrate the calculation.
The paper also includes a proposed approach, the tw o mile method, to quantify engineering
productivity loss using data fo r similar work, but different complexity.

CDR.980.1
 2012 AACE INTERNATIONAL TRANSACTIONS

Table of Contents

Abstract........................................................................................................................................... 1

List of Tables.................................................................................................................................. 3

List of Figures................................................................................................................................. 3

List of Equations............................................................................................................................. 3

In tro d u ctio n ................................................................................................................................... 4

Measured Mile/Baseline M ethod andits Application in C onstruction.............................. 5

Engineering Productivity M easurem ent.................................................................................... 6

Measured Mile/Baseline M ethod to Calculate Lost Engineering P ro du ctivity..................... 8

Case Study.....................................................................................................................................10

Two Mile M ethod to Q uantify Engineering Productivity Loss

w ith Complexity D ifference....................................................................................................... 12

Conclusion.....................................................................................................................................16

References................................................................................................................................... 17

CDR.980.2
 2012 AACE INTERNATIONAL TRANSACTIONS

List of Tables
Table 1, Piping Productivity Data.............................................................................................11

Table 2, Piping Design Productivity Baseline......................................................................... 11

Table 3, Piping Design Productivity Loss.................................................................................12

List of Figures

Figure 1
Correlation between W ork Hours and Number of Lines in the Baseline.......................... 13

List of Equations

Equation 1 .................................................................................................................................... 14

Equation 2 .................................................................................................................................... 14

Equation 3 .................................................................................................................................... 14

Equation 4 .................................................................................................................................... 14

Equation 5 .................................................................................................................................... 14

Equation 6 .................................................................................................................................... 14

Equation 7 .................................................................................................................................... 14

Equation 8 .....................................................................................................................................15

Equation 9 .....................................................................................................................................15

CDR.980.3
 2012 AACE INTERNATIONAL TRANSACTIONS

Introduction
Proving or quantifying lost productivity is one o f the most contentious and controversial areas in
construction claims and disputes. It is a task that usually involves processing data fo r completed
w ork and corresponding w ork hours, calculating productivity, establishing baseline, causation
study, and measuring productivity loss.

Although there may exist various methods to quantify productivity loss, the measured
mile/baseline method by far is the most widely accepted fo r com puting lost productivity in the
construction industry. The original measured mile method determines the benchmark of the
productivity achieved on identical activities in the same project. The productivity achieved on
unimpacted sections was compared to the productivity achieved in sections where the work is
known to be impacted. Responsibility fo r the difference in productivity is then assigned to the
impacting party. Since it is not always possible to segregate an unimpacted area o f work, the
measured mile calculation has subsequently broadened to a baseline method. The baseline method
also compares similar work activities, yet it allows fo r the comparison of lightly impacted areas to
more heavily impacted areas. The lightly impacted portion is defined as the baseline set and
although the baseline productivity rate may still contain lost productivity, it would allow the
claimant to measure the productivity loss in the heavily impacted areas, and assign that loss to the
responsible parties. In this paper, the authors use the broad meaning of "measured mile m ethod,"
and do not differentiate it from the baseline method.

During the course o f design and construction, changes and the ripple effect of m ultiple changes are
one of the most im portant causes fo r engineering productivity loss. It is usually difficult to
discretely quantify the productivity loss caused by cumulative impacts o f m ultiple changes. Despite
wide application and acceptance in proving labor productivity loss, the measured mile/baseline
method has not been successfully applied to engineering productivity.

The challenges lie in engineering productivity measurement and the determ ination of project
specific productivity benchmarks. In this paper, the procedures specifically fo r measured
mile/baseline calculation o f engineering productivity are developed and the im plem entation is
demonstrated by a case study, which proves that engineering productivity loss can be successfully
quantified if the proposed procedures are followed to correctly measure productivity and
determ ine the measured mile or baseline.

To overcome the difficulties of finding engineering w ork at exactly the same com plexity in a
project, a broader approach, called the tw o mile method, is proposed to calculate engineering
productivity loss based on the data from similar areas (segments, systems, components, units ...)
w ith different com plexity by establishing a linear relationship between productivity and complexity.
This paper not only contributes to the literature on engineering productivity loss calculation, also
its practical significance is substantiated by offering viable approaches to quantify engineering
productivity loss in litigation proceedings. Although this paper only focuses on the quantification of
engineering productivity loss, the importance of cause and effect analysis, and responsibility
allocation should not be ignored.

CDR.980.4
 2012 AACE INTERNATIONAL TRANSACTIONS

Productivity can be measured as the ratio between the units of work accomplished (output) and
the units o f tim e or efforts expended (input). In the literature, tw o inverse definitions were used:
in put/ou tp ut, and o u tp u t/in p u t. In this paper, productivity is defined as the ratio o f input to
output, but the choice is a m atter of personal preference.

Measured Mile/Baseline Method and its Application in Construction

Although the measured mile method has been widely used since 1986, the core fundam ental of
measured mile analysis originated much before that. In 1951, the measured mile method was
successfully used in Maryland Sanitary Mfg. Co. v. US [20]. Researchers and professionals have
developed and proposed systematic procedures fo r measured mile or baseline calculation [6, 12,
21, 22]. The general steps to perform measured mile/baseline calculation include processing data
fo r production input and output, identification o f units representing the completed work,
com puting the productivity by tim e periods or w ork sections, identification o f measured mile or
baseline, and using the difference between the impacted productivity and the measured
mile/baseline to calculate productivity loss and damages.

A systematic procedure fo r measured mile calculation as a method was introduced by Zink [21, 22].
The measured mile method compares the productivity in an umimpacted period of tim e w ith
productivity in an impacted period o f tim e in the same project. The procedures Zink proposed
include the following.

• Plot the actual labor worker-hours expended versus corresponding percentage of

com pletion fo r the w ork;
• Exclude the first and last 10% from the analysis; and,
• Identify a linear or near linear portion showing the most efficient rate of progress in the
80% of the curve as the measured mile.

The measured mile method, as originally proposed, requires that the portion of w ork used to
calculate the measured mile must be free or essentially free of disruptions and continuous in tim e,
which limits its application. In order to address the weaknesses in the measured mile method,
Thomas and his collaborators introduced the baseline concept [12, 13, 15, 16, 17, 18, 19]. Thomas
defined that a baseline period is a period of tim e when the best productivity is achieved, but its
baseline procedure requires the baseline to be selected from the reporting periods w ith highest
production [19]. The baseline period is not required to a continuous, unimpacted tim e frame,
while the measured mile period has to be a consecutive set of tim e periods. The steps to
determ ine a baseline proposed by Thomas and his collaborators can be summarized as follows.

• Determine the total num ber of reporting periods;

• The size of the baseline subset, n, is 10% o f the total num ber of reporting periods and
should not be less than 5;
• The contents of the baseline subset are the n reporting periods th a t have the highest
production or output; and,
• The baseline is the median of productivity value in each reporting period or the quotient of
total production divided by total input in the baseline subset.

CDR.980.5
 2012 AACE INTERNATIONAL TRANSACTIONS

In the attem pt to overcome the shortcomings of having subjective percentage used in selecting the
measured mile or baseline, Ibbs and Liu proposed a statistical clustering method to separate the
productivity data into different groups [6]. The average value fo r the group w ith the best
productivity would serve as the measured mile or baseline. Although Ibbs and Liu's method could
select the best baseline based on the available data, and was claimed to be more objective and
neutral, the baseline selected by Ibbs and Liu's method may include data points that does not
represent the best-sustained productivity.

Zink's measured mile method requires that the measured mile needs to be in a period w ith a linear
or near linear relationship between the expended labor hours and percent of w ork complete,
which implies that the level of productivity should maintain in the period that is disruption free or
nearly so. The assumption behind the baseline method proposed by Thomas fo r selecting the
reporting periods w ith the highest production or output as the baseline period can be understood
as that the disruption impact per output unit is relatively lower in the periods in which best
production is achieved. Ibbs and Liu's method employs a statistical clustering technique to
segregate the periods w ith best productivity from all the others and uses the average as the
baseline productivity, which implies that the baseline period is not necessarily continuous. In terms
of baseline selection, Ibbs and Liu's method may include unstained productivity data point into the
baseline set, because no productivity o f a reporting period outside the baseline period can be
better than that in the baseline period when relying on this technique. Zink and Thomas's
procedures appear to seek the best-sustained productivity, and as a result, the reporting period
w ith the highest productivity if not sustained would appear outside the measured mile/baseline
set.

Despite the differences among various procedures, the core of measured mile/baseline method is a
comparison of productivity in an unimpacted or lightly impacted area w ith the productivity of
similar w ork in an impacted area. The comparison relies on an unbiased baseline representing the
best sustained productivity.

Engineering Productivity Measurement

Reliable labor productivity measurement is a crucial element of measured mile/baseline method to
prove productivity loss in construction, and then reliable engineering productivity measurement is
also crucial to the method's application to proving engineering productivity loss. In spite of the
awareness of the essentiality of engineering productivity, there have been only limited studies in
engineering productivity measurement.

Thomas and his co-authors proposed a conceptual model fo r measuring productivity on design and
engineering, in which conversion factors and rule of credit were used to account fo r the differences
of various work products in different stage of design engineering [14]. The application of this
conceptual model is discouraged by the involvement of conversion factors and rule of credit, which
are heavily affected by personal preference. The accuracy o f the model is also compromised due to
lack of an industry consensus on design document complexity.

CDR.980.6
 2012 AACE INTERNATIONAL TRANSACTIONS

The Engineering Productivity Measurement Research Team of Construction Industry Institue (CII),
or Research Team 156 (RT 156), acknowledged that there was no standard measurement of
productivity in the engineering phase fo r internal and external benchmarking [4]. CII RT 156
proposed to use measurable installed quantities fo r the measurement of design output. RT 156
conducted a detailed evaluation of the piping design process and found a good correlation
between the engineering productivity measured as worker-hours spent per linear fo o t of piping
and the number of equipm ent pieces in the piping systems. On the basis that engineering
productivity is affected by project complexity, quality of input, quality o f design output, RT 156
proposed a conceptual model fo r engineering productivity, in which engineering productivity is
measured as engineering hours per installed quantity adjusted by three factors, input quality
factor, scope and com plexity factors, and design effectiveness factor.

Song and others focused on the activities of steel drafting to characterize com plexity factors for
engineering productivity through literature review, interview and survey [10]. They proposed a
neutral network model o f engineering productivity quantification to facilitate project management
using 17 project-level influence factors, including project type, contract type, piece cloning, etc. In
contrast to project level factors, Song and AbouRizk proposed a conceptual scope definition
method, quantitative engineering project scope definition (QEPSD), to measure engineering project
scope quantitatively and quantify engineering com plexity at the element level, such as number of
fittings and drafted quantity fo r steel drafting [11]. In QEPSD, different design components are
converted into the same design units using com plexity functions, which may only exist
conceptually.

For the purpose of establishing a precise and reliable engineering productivity benchmarking in the
construction industry, the CII Benchmarking and Metrics (BM &M ) program developed the
Engineering Productivity M etric System (EPMS), which included quantity-based metrics. BM&M
proposed using direct work-hours as the input and issued fo r construction (IFC) quantities as the
output to measure engineering productivity [7]. The EPMS includes a set of metrics hierarchically
at m ultiple levels, including Level I (project), Level II (major category), Level III (sub-category), and
Level IV (element). EPMS consists o f six major categories, including concrete, steel, electrical,
piping, instrum entation, and equipm ent. Engineering productivity was proposed to be measured
at various levels and categories. BM&M collected engineering productivity data from 87 projects
provided by 21 different ow ner and contractor companies and examined the engineering
productivity in different categories and levels. BM&M also performed a correlation analysis
between direct engineering work-hours and IFC design quantities on structural steel and piping,
and found statistically significant bivariate relationships in those tw o categories. CII's research
e ffo rt also included the method to calculate a comprehensive index fo r productivity assessment at
the project level based on the interdependencies among various disciplines and at different levels
[8].

Other researches on engineering productivity include Chang and Ibbs and Ebrahimy and Rokni [2,
5]. Chang and Ibbs used a deliverable-based engineering productivity measure. They examined
various engineering productivity factors, and highlighted tw o significant ones: phase involvement
and project size. Ebrahimy and Rokni illustrated the expected values fo r existing metrics in the
industry are dependent on the size o f the projects.
CDR.980.7
 2012 AACE INTERNATIONAL TRANSACTIONS

In summary, although various attem pts have been made to establish a standard measurement of
engineering productivity and develop engineering productivity benchmarking fo r construction
industry, the models and methods proposed by various researchers remain in a conceptual stage.
Moreover, researchers still differ significantly in how to quantify the measurement of engineering
output, but most researchers agree that engineering productivity is influenced by m ultiple factors,
including project size and complexity, which could explain why a linear or close to linear
relationship was not observed between engineering work-hours and output quantities based on
the historical data from various projects.

Measured Mile/Baseline Method to Calculate Lost Engineering Productivity

Although a consensus on reliable engineering productivity measurement fo r construction industry
has not been widely reached, it does not prevent the measured mile/baseline method from being
an effective method to calculate engineering productivity loss. First, measured mile/baseline
method has been a widely accepted method in the construction industry. Second, measured
mile/baseline method is project specific, which is not influenced by the factors that differ from
project to project.

In the measured mile/baseline method fo r labor productivity, the measured mile/baseline is

defined as the period of tim e during which the labor productivity was unimpacted or lightly
impacted. This definition is not appropriate fo r engineering work. According to AIA Document
B141 [1], design professionals provide services in five phases, schematic design, design
development, construction document, bidding or negotiation, and construction or contract
adm inistration. For a design build project, the design and engineering work usually cover from
design development, construction documents through the end of construction. The output for
design professionals during the period from design development through the end of construction
includes drawings, specifications, calculations, data sheets and others. The engineering efforts are
also spent in both design and construction phases. The engineering output fo r a particular item in
a project may go through m ultiple design revisions as the construction proceeds. Therefore, it is
more appropriate to define the engineering measured mile/baseline as the area (segment, system,
component, unit ...) where engineering productivity was not impacted or lightly impacted.

Loulakis and Santiago presented the follow ing basic guidelines that should be followed in
determ ining a measured mile fo r labor productivity [9].

• The work perform ed in the measured mile and the impacted period should be substantially
similar in type, nature and complexity;
• The composition and level o f skill of crews should be comparable; and,
• The above principles should also apply to the measured mile/baseline fo r engineering
productivity.

The output used fo r engineering productivity calculation needs to represent the scope of
construction work that is designed and engineered. There are three different ways to quantify
engineering output in literature:

CDR.980.8
 2012 AACE INTERNATIONAL TRANSACTIONS

Quantities of Engineering Documents [2][14]

In order to account fo r the differences among various documents, conversion is needed to
normalize the quantities, which compromises the reliability of this measurement because of the
lack o f a standard definition of com plexity and content, and the use o f CAD technology [3, 4, 11].
Although this approach is not preferred, the authors would not rule out the possibility that under
some special circumstances, it may be appropriate to quantify engineering output w ith the mutual
agreements between the parties in the project.

Installed Quantities or the IFC Quantities

CII used the installed quantities and the IFC quantities to measure engineering output. CII's
research indicates that significant correlation exists between expended engineering work-hours
and engineering output measured w ith Installed quantities or the IFC quantities [4, 7].
Other Quantifiable Elements for the Scope of Work
Song and others proposed to use the quantity of design items w ithin a discipline, such as a beam or
a w indow to measure engineering output [11].

Since this paper is focused on a general measured mile/baseline method fo r engineering

productivity, engineering output quantification is not specified here, and it should be chosen
depending on the data and the characteristics o f the w ork to be designed and engineered. The
measured mile/baseline calculation fo r engineering productivity can be performed using the
follow ing procedure:

• Segregate the scope o f engineering work into areas (segments, systems, components, units
...).
• Q uantify work hours fo r each area (segment, system, component, unit ...).
• Q uantify engineering output fo r each area (segment, system, component, unit ...).
• Calculate engineering productivity fo r each area (segment, system, component, unit ...).
• Identify unimpacted or lightly impacted areas (segments, systems, components, units ...) as
the measured mile/baseline subset. (This is generally determined by the magnitude of
changes, modifications, or review oversight initiated by the ow ner or third parties, and the
engineering e ffo rt required to address them)
• The measured mile/baseline productivity is the productivity of the measured mile/baseline
subset using total work-hours divided by the total quantities representing design output in
the same subset.

It is known that disruptions reduce productivity; therefore, it is natural to rule out disrupted areas
(segments, systems, components, units . ). The proposed method is only dependent on the
causation study to determ ine the baseline, which is different from Zink's measured mile procedure,
Thomas's baseline method, and Ibbs and Liu's K-means clustering technique based approach. In
order to test the quality of the baseline, a correlation analysis on the measured mile/baseline set
may be performed. Furthermore, the above baseline selection and calculation should be
corroborated by a cause-and-effect analysis to ascertain the causes fo r the productivity losses, and
allocate the losses to the responsible parties.

CDR.980.9
 2012 AACE INTERNATIONAL TRANSACTIONS

Case Study
The proposed measured mile/baseline method to calculate engineering productivity loss has been
applied to a m ulti-hundred m illion dollar international project to design and build a process plant.
The project experienced m ultiple ow ner responsible impacts, which caused trem endous cost
overrun. The ow ner issued extensive change orders and the EPC contractor was paid fo r part of
the direct cost of the changes estimated fo r the normal conditions, adding up to several hundred
m illion dollars.

There was, however, still millions of dollars in engineering that remained uncompensated, which
could not be discretely captured in those change orders. During the course of the design and
construction of the project, the ow ner and its representatives issued tens o f thousands of design
comments, the m ajority of which were ow ner changes. The measured mile/baseline method was
used to quantify the engineering productivity loss, which was impossible to be captured discretely
in the change orders, because of the cumulative impacts o f numerous changes that the ow ner was
responsible for. This case study is limited to the piping discipline to dem onstrate the application of
the proposed method. For confidentiality reason, productivity data used here was scaled and
altered from the actual.

The engineering w ork on piping is divided by various systems in the case study and the number of
pipe lines is used to measure engineering output. Detailed project cost reports were evaluated to
segregate and summarize direct work-hours on each system. Then the as built P&ID and isometrics
were studied to identify the number of pipe lines in each system. The engineering productivity in
this case study was defined as the work-hours spent on a pipe line. The direct work-hours, number
of pipe lines, and calculated productivities are summarized in table 1.

CDR.980.10
 2012 AACE INTERNATIONAL TRANSACTIONS

System Name ID Direct Hours Number of Hours Per Line

Lines
Boiler Blowdown 140 5 28.0
Boiler Feedwater 340 7 48.6
Chemical A 20 1 20.0
Chemical B 730 24 30.4
Chemical C 190 2 95.0
Chemical D 10 1 10.0
Fire Protection Foam 190 4 47.5
Fire W ater 26,870 144 186.6
Instrum ent Air 1,370 95 14.4
Liquid State Chemical B 19,690 90 218.8
Low Pressure Steam 7,550 350 21.6
Lube Oil 50 2 25.0
Process W ater 1,000 44 22.7
Service Air 960 55 17.5
Service W ater 7,005 275 25.5
Steam Condensate 7,600 330 23.0
Tempered W ater 365 17 21.5
Waste W ater 1,650 29 56.9
Table 1— Piping Productivity Data

According to the contract, the design builder was required to prepare a 90% design, and then
complete the IFC design upon the receipt of the owner's comments on 90% design. In this case
study, systems were segregated into those where the ow ner made few er or no changes, and those
where the ow ner made significant changes. The systems summarized in table 2 are those w ith
little or no ow ner changes, so they were selected as the measured mile/baseline. The baseline
productivity was calculated to be 22.2 hours per line.

System Name Direct Hours Number Hours Per Line

of Lines
Boiler Blowdown 140 5 24.0
Chemical A 20 1 16.0
Chemical D 10 1 12.0
Instrum ent Air 1,370 95 14.4
Low Pressure Steam 7,550 350 22.3
Lube Oil 50 2 25.5
Process W ater 1,000 44 23.0
Service Air 960 55 17.4
Service W ater 7,005 275 25.1
Steam Condensate 7,600 330 23.0
Tempered W ater 365 17 20.6
Total 26,070 1,175 22.2
Table 2— Piping Design Productivity Baseline

CDR.980.11
 2012 AACE INTERNATIONAL TRANSACTIONS

The remaining systems were determined to be the impacted, as either 90% design had to be
redone due to m odification by the ow ner or the IFC documents were revised m ultiple times
because o f the owner's changes. The productivity loss calculation is summarized in table 3.

In order to assess the quality of the measured mile/baseline subset, a correlation analysis between
num ber of lines and work-hours expended is conducted fo r the subset, as shown in figure 1. The
calculated correlation coefficient was 0.9869, which shows a significant linear relationship between
the number of lines and work-hours expended in the case study. This indicates that in the case
study, not only the number of lines is a good m etric to quantify piping design and engineering, but
also the selected measured mile/baseline has been proven to be consistently unimpacted or lightly
impacted systems.

System Name Direct Number Hours Per Baseline Productivity

Hours of Lines Line Loss
Boiler Feedwater 340 7 48.0 22.2 181
Chemical B 730 24 30.3 22.2 195
Chemical C 190 2 92.5 22.2 141
Fire Protection Foam 190 4 47.5 22.2 101
Fire W ater 26,870 144 186.6 22.2 23,670
Liquid Chemical B 19,690 90 218.8 22.2 17,692
Waste W ater 1,650 29 56.9 22.2 1,007

Table 3— Piping Design Productivity Loss

CDR.980.12
 2012 AACE INTERNATIONAL TRANSACTIONS

Figure 1 — Correlation Between Work Hours and Number of Lines in the Baseline

Two Mile Method to Quantify Engineering Productivity Loss with Complexity Difference

The underlying premise fo r the measured mile/baseline method is comparison o f identical or

similar work w ith the same complexity. Unlike labor productivity, which can be isolated and
sampled daily, weekly or m onthly, sampling engineering productivity requires the whole life cycle
of design and construction. The engineering e ffo rt on a particular area (segment, system,
component, unit ...) o f the facility to be built does not end when the IFC documents are issued;
instead the e ffo rt may continue until the construction is completed. Furthermore, it is impractical
to record engineering work-hours fo r each installed unit of work.

Engineering productivity sampling relies on the work break down in the project cost accounting
system, and the tim e keeping efforts by every designer/engineer on the project. Even though it
may be possible to measure engineering productivity fo r each dividable area (segment, system,
component, or unit . ) under a well designed project cost accounting system and w ith the diligent
efforts fo r each engineering tim e keeper, the number of comparable areas (segments, systems,
components, or units ...) are often limited, and finding quantifiable engineering w ork w ith the same
com plexity is still very difficult. Therefore, although the concept o f measured mile/baseline
method is straightforw ard, the im plem entation in practice may be challenging because o f the
difficulties in finding the measured mile/baseline w ith the same com plexity as the impacted.

RT 156 o f CII researched the relationship between piping engineering w ork hours and physical
measurement of quantities, and concluded that equipm ent piece count represents the complexity
of piping engineering [4]. RT 156 then form ulated a linear relationship between the piping

CDR.980.13
 2012 AACE INTERNATIONAL TRANSACTIONS

engineering productivity measured as work-hours per installed quantity and number of pieces of
equipm ent in the piping discipline based on a regression analysis on the data from 40 projects.

Hours/Feet of Piping = 0.44985 + 0.00134 • (Number of Pieces of Equipment)

equation 1

The above finding by RT 156 o f CII indicates that piping engineering productivity when measured as
work hours per feet o f piping is driven by the number o f pieces o f equipm ent, and the more pieces
of equipm ent that are in the piping system, the more engineering hours are expended. This finding
proved that it is a practical and reasonable assumption to model engineering productivity (P) as a
linear function of a variable (C) that represents engineering com plexity (complexity variable), or

P = a •C + b
equation 2
P = engineering productivity.
C = the variable representing engineering complexity.
a, b = parameters.

If productivity and com plexity inform ation are available fo r tw o not impacted (or lightly impacted)
areas (segments, systems, components, or units ...), or

P1 = a • C 1 + b
equation 3

P2 = a • C 2 + b
equation 4

then parameters a and b can be solved as

P
1 1
- P
1 2
a =
C 1 - 2C
equation 5

C P - ^C 2 P
^'11 2 M
b =
C 1 - 2C
equation 6

Given the actual productivity P3,a per linear fo o t o f piping and com plexity C3 fo r an impacted area
(segment, system, component, or unit ...), the productivity that would have been w ith o u t impacts
(or lightly impacted) P3 is

P3 = a • C 3 + b
equation 7
CDR.980.14
 2012 AACE INTERNATIONAL TRANSACTIONS

or

P - P CP - C P
P3 = -pL - p L . C 3 + ^ 11 2 '-'21 1
C - C
'“'1 ^2 C 1 - 2C
equation 8

Then the productivity loss on a linear fo o t o f piping fo r the impact area is

AP = P - P
equation 9

Since data from tw o unaffected (or lightly impacted) areas (segments, systems, components, or
units ...) is needed to solve equations (3) and (4) fo r parameters a and b, this method is named the
tw o mile method. If there are more than tw o unaffected areas (segments, systems, components,
or units ...) available in the unaffected set, or the measured mile/baseline set, least square
regression is recommended to approximate the parameters a and b th a t would be representative.
In fact, the size o f the measured mile/baseline set should not be limited to 2, and it should include
all the areas (segments, systems, components, or units . ) that can be identified as unaffected or
lightly impacted. The advantage o f this method is th a t it makes it possible to calculate engineering
productivity at different com plexity w ith as few as tw o measured mile/baseline needed. The
procedures to im plem ent the tw o mile method, summarized below, are similar to the measured
mile/baseline method, and are also very straightforw ard:

• Segregate the scope o f engineering work into areas (segments, systems, components, units
...).
• Q uantify work hours fo r each area (segment, system, component, unit ...).
• Q uantify engineering output fo r each area (segment, system, component, unit ...).
• Calculate engineering productivity fo r each area (segment, system, component, unit ...).
• Identify the com plexity factor, and determ ine the com plexity value fo r area (segment,
system, component, unit ...).
• Determine all the unaffected or lightly impacted areas (segments, systems, components,
units . ) as the benchmark.
• If there are only tw o areas (segments, systems, components, units ...) in the benchmark
subset, go to 8; if there are more than tw o areas (segments, systems, components, units ...),
go to 9.
• Use Equations (5) and (6) to calculate parameters a and b, go to 10.
• Use least square regression to determ ine parameters a and b.
• Use Equations (8) and (9) to calculate productivity loss fo r a linear fo o t of piping in the
impacted area (segment, system, component, unit ...). And,
• The total productivity loss fo r a given area (segment, system, component, unit ...) is
calculated as the product o f the productivity loss per linear fo o t and the total pipe length in
the area (segment, system, component, unit ...).

CDR.980.15
 2012 AACE INTERNATIONAL TRANSACTIONS

Conclusion
This paper has reviewed and summarized the measurement technique o f engineering productivity,
and applied the measured mile/baseline method to quantify engineering productivity loss fo r
cumulative impacts o f m ultiple disruptions, which are difficult to quantify discretely. The
im plem entation has been demonstrated by a case study, which illustrates that project specific
inform ation can be successfully used to prove engineering productivity loss. In addition, in order to
alleviate the difficulty o f identify measured miles or baseline from areas (segments, systems,
components, units ...) w ith same complexity, a tw o mile method is proposed. The tw o mile method
is designed to calculate engineering productivity loss fo r similar w ork w ith different com plexity
under the condition that engineering productivity is a linear function of engineering complexity.
The proposed method is straightforw ard and easy to understand and use.

CDR.980.16
 2012 AACE INTERNATIONAL TRANSACTIONS

1. American Institute o f Architects (AIA) Standard Form of Agreement Between Owner and
Architect, AIA Document B141, AIA, Washington, DC, 1997.
2. Chang, A.S., and W. Ibbs. S ystem M o d e l f o r A n a ly z in g D esign P ro d u c tiv ity , Journal of
Management in Engineering, 22, 1 (2006): Pgs. 27-34. ASCE, Reston, VA.
3. Chang, L., M.E. Georgy, and L. Zhang. Engineering Productivity Measurement, Research
Report 156-11 CII, Austin, TX, 2001.
4. Construction Industry Institute (CII). Engineering Productivity Measurement, CII Publication
156, CII, Austin, TX, 2001.
5. Ebrahimy, Y., and S. Rokni. V a lid ity o f In d u s try B enchm arks a n d M e tric s f o r E ng in e ering
P ro d u c tiv ity . Proceedings, Construction Research Congress, Banff, Canada, 2010.
6. Ibbs, W., and M. Liu. Im p ro v e d M e a s u re d M ile A na lysis Technique. Journal of Construction
Engineering and Management, 131, 12 (2005): Pgs 1249-1256. ASCE, Reston, VA.
7. Kim, I. Development and Implementation of an Engineering Productivity Measurement
System (EPMS) for Benchmarking. Ph.D. thesis, University of Texas, Austin, TX, 2007.
8. Liao, P.C. 2008 Influence Factors of Engineering Productivity and Their Impact on Project
Performance, Ph.D. thesis, University of Texas, Austin, TX.
9. Loulakis, M.C., and S.J. Santiago. G e ttin g th e M o s t O u t o f Your "M e a s u re d M ile " A p p ro a ch ,
Civil Engineering, 69, 11 (1999): Pgs. 69. ASCE, Reston, VA.
10. Song, L., M. Allouche, and S.M. AbouRizk. M e a s u rin g a n d E s tim a tin g D ra ftin g P ro d u c tiv ity ,
Proceedings. Construction Research Congress, Honolulu, HI, 2003.
11. Song, L., S.M. AbouRizk. Q u a n tify in g E n g in e ering P ro je c t Scope f o r P ro d u c tiv ity M o d e lin g .
Journal of Construction Engineering and Management, 131, 3 (2005): Pgs. 360-367. ASCE,
Reston, VA.
12. Thomas, H.R., and I. Zavrski. C o n stru ctio n B aseline P ro d u c tiv ity : T heory a n d P ra ctice Journal
of Construction Engineering and Management, 125, 5 (1999): Pgs. 295-303. ASCE, Reston,
VA.
13. Thomas, H.R., D.R. Riley, and V.E. Sanvido. Loss o f L a b o r P ro d u c tiv ity Due to D e live ry
M e th o d s a n d W e a th e r. Journal of Construction Engineering and Management, 125, 1
(1999): Pgs. 39-46. ASCE, Reston, VA.
14. Thomas, H.R., Q.C. Korte, V.E. Sanvido, and M.K. Parfitt. C o n ce p tu a l M o d e l f o r M e a s u rin g
P ro d u c tiv ity o f D esign a n d E ngineering. Journal of Architectural Engineering, 5, 1 (1999):
Pgs. 1-7. ASCE, Reston, VA.
15. Thomas, H.R., and V.E. Sanvido. Q u a n tific a tio n o f Losses Caused b y L a b o r Ine fficie ncies
W h e re is th e Elusive M e a s u re d M ile ? Construction Law and Business, 1, 3 (2000): Pgs. 1-14.
Aspen Law & Business, New York, NY.
16. Thomas, H.R., and V.E. Sanvido. Role o f th e F a b ric a to r in L a b o r P ro d u c tiv ity . Journal of
Construction Engineering and Management, 126, 5 (2000): Pgs. 358-365. ASCE, Reston, VA.
17. Thomas, H.R., M.J. Horman, R.E. Minchin, and D. Chen. Im p ro v in g L a b o r F lo w R e lia b ility f o r
B e tte r P ro d u c tiv ity as Lean C o n stru ctio n P rinciple. Journal of Construction Engineering and
Management, 129, 3 (2003): Pgs. 251-261. ASCE, Reston, VA.
18. Thomas, H.R., R.E. Minchin, and D. Chen. Im p ro v in g Role o f W o rk fo rc e M a n a g e m e n t in
B rid g e S u p e rs tru c tu re L a b o r P ro d u c tiv ity . Journal of Management in Engineering, 19, 1
(2003): Pgs. 9-16. ASCE, Reston, VA.

CDR.980.17
 2012 AACE INTERNATIONAL TRANSACTIONS

19. Thomas, H.R. Q u a n tific a tio n o f Econom ics Losses Caused b y L a b o r Inefficiencies.
Proceedings. Construction Super Conference, San Francisco, CA, 2007.
20. Wickwire, J.M., T.J. Driscoll, and S.B. Hurlbut. C h a p te r 10, C a lcu la tin g C o n tra c to r's D am ages,
Construction Scheduling: Preparation, Liability, and Claims Implementing, First Edition,
(1991):page 302, John W iley & Sons, Inc., New York.
21. Zink, D.A. Im p a c ts a n d C o n stru ctio n In e fficie n cy. AACE International Cost Engineering
Journal, 32, 11 (1990): Pgs. 21-23. AACE International, M organtown, WV.
22. Zink, D.A., The M e a s u re d M ile : P ro vin g C o n stru ctio n In e ffic ie n c y Costs. AACE International
Cost Engineering journal, 28, 4 (1986): Pgs. 19-21. AACE International, M organtown, WV.

Dr. Tong Zhao, PE PSP

Delta Consulting Group, Inc.
[email protected]

J. Mark Dungan
Delta Consulting Group, Inc.
[email protected]

CDR.980.18