CAUSES O F QUALITY DEVIATIONS IN DESIGN

AND CONSTRUCTION
By James L. Burati Jr., 1 Member, ASCE, Jodi J. Farrington, 2 Associate
Member, ASCE, and William B. Ledbetter, 3 Fellow, ASCE
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

ABSTRACT: Quality deviation data are collected from nine fast-track industrial
construction projects. The data are collected after the construction phase of the
projects and identify the direct costs associated with rework (including redesign),
repair, and replacement. Analyses of the data indicate that deviations on the proj-
ects accounted for an average of 12.4% of the total project costs. Furthermore,
design deviations average 78% of the total number of deviations, 79% of the total
deviation costs, and 9.5% of the total project cost. Construction deviations average
16% of the total number of deviations, 17% of the total deviation costs, and 2.5%
of the total project cost. These values are conservative because they consider only
direct costs, but they are indications of the areas that cause the greatest impact on
total project costs. By applying the procedures presented in this paper to its his-
torical records, a company can identify not only its deviation costs, but also the
most likely areas on which to focus to reduce these costs on future projects.

BACKGROUND
In recent years, participants in the construction industry have become
increasingly aware of rising construction costs and perceptions of increased
quality problems. Business Week ("Quality" 1982) stated that manufacturers
agreed that 15-20% of their sales dollars were used to correct mistakes,
and that the way to increase profit was to reduce the cost of poor quality
rather than to increase sales. At a national conference on quality assurance
in the building community, it was suggested that the cost of poor quality
was at least 7.5% of the value of new nonresidential work, and the question
was raised as to why the construction industry did not follow the lead of
the manufacturing industry, and increase profit by reducing the costs of
quality problems (Shilstone 1983).
Research performed in conjunction with the Business Roundtable's con-
struction industry cost-effectiveness project revealed that on-site construc-
tion activities for industrial projects were perceived to be the most troubling
quality problems (Ledbetter 1983). The troublesome items were identified
as concrete, piping, welding, roofing, painting, and electrical work. In ad-
dition, investigation of the frequency and severity of claims on federally
funded and administered projects found that design errors were the single
most common cause for contract claims, accounting for 46% of the additive
claims that were reviewed (Diekmann and Nelson 1985).
These results appear to be in conflict, with one indicating that on-site
construction is the most important source of quality problems and the other
indicating that design errors are the major cause of increased costs. Ques-
tions regarding the causes, number, and costs of quality problems in design
and construction remained unanswered.
'Prof, of Civ. Engrg., Clemson Univ., Clemson, SC 29634-0911.
2
Asst. Mgr. Tech. Services, Nat. Industrial Constructors, Birmingham, A L 35243.
3
Quality Mgmt. Consultant, 1006 Challedon Way, Pendleton, SC 29670.
Note. Discussion open until August 1,1992. To extend the closing date one month,
a written request must be filed with the ASCE Manager of Journals. The manuscript
for this paper was submitted for review and possible publication on January 8, 1990.
This paper is part of the Journal of Construction Engineering and Management, Vol.
118, No. 1, March, 1992. © A S C E , ISSN 0733-9364/92/0001-0034/$1.00 + $.15 per
page. Paper No. 26562.

34

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 RESEARCH OBJECTIVE

The research presented in this paper was conducted to identify the causes
and magnitude of quality problems in design and construction and to de-
termine the costs associated with the quality problems. The research was
conducted under the guidance of the quality management task force (QMTF)
of the Construction Industry Institute (CII).
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

RESEARCH SCOPE AND METHODOLOGY

The objectives were met through the use of interviews, both in person
and by telephone, visits to home offices and construction sites, and analysis
of data supplied by cooperating firms. A glossary of definitions of key quality
terms was developed to ensure consistency in the collection and categorizing
of data (see Appendix I). To ensure meaningful comparisons, data collection
was limited to industrial projects for which construction began before project
design was completed. Data from nine projects were then summarized and
categorized to identify the causes and numbers of quality problems along
with their associated costs.

GLOSSARY OF QUALITY-RELATED TERMS

The QMTF developed a glossary of quality terminology that was not

industry-specific, but was intended to have wide applicability in the con-
struction industry. The QMTF Glossary of Quality Terminology is presented
in Appendix I. A complete discussion of the glossary and its development
is presented in (Burati and Farrington 1987). Two key definitions used in
the glossary are for the terms quality and deviation.
Quality is defined as "conformance to established requirements." By
avoiding dealing with degree of goodness or satisfaction, this definition
provides a basis for measurement, i.e, the requirements are either met or
not met. This definition is essential to the objective of identifying costs
associated with quality problems.
The term deviation, rather than failure or defect (which are commonly
used in manufacturing industries), indicates that a product or result that
does not fully conform to all specification requirements does not necessarily
constitute an outright failure (Davis et al. 1989). Deviation includes changes
to the requirements that result in rework, as well as products or results that
do not conform to all specification requirements, but do not require rework.

DATA COLLECTION PROCEDURES

All CII member firms were surveyed to identify those willing to provide
projects on which quality deviation data could be collected. To allow for
meaningful comparisons, project selection was limited to industrial projects
of at least $5,000,000 in total installed costs that were completed in the
mid-1980s and that each had a different designer and constructor. Table 1
presents brief descriptions for the nine projects that were studied in detail.

Types of Construction
The projects were classified according to the following types:

• New construction: the design and construction of the structure to house

35

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 the operation as well as the installation of the process equipment and
instrumentation.
• Retrofit construction: the design and installation of process equipment
and instrumentation in an existing, but empty, structure. Removal of
existing process systems prior to installation of new systems was not
part of the retrofit projects considered in the study.
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

e
Upgrade construction: the design and construction associated with the
modification of an existing facility, for example, technological advance-
ments and process improvements.

COLLECTION OF DATA

Wherever possible, data were collected directly from field change orders,
requests for information, or design change orders that contained complete
information concerning (1) A description of the change; (2) why the change
was required; (3) who initiated the change; and (4) the cost of the change.
When this information was not available, other sources, including cost ac-
counting tabulations and computerized project summaries, were investi-
gated. When complete information was not available from the alternate data
sources, additional information was obtained through interviews with proj-
ect representatives in order to gain sufficient understanding for analysis of
the data.

DEVIATION DATA COLLECTED

The deviation data that were collected and analyzed were limited to the
design and construction (including fabrication) phases of the projects stud-
ied. The deviation data collected included only the direct costs associated
with rework (including repair and replacement) and therefore do not con-
stitute the total costs associated with the deviations. These direct costs of
correcting deviations are only the "tip of the iceberg." Impact costs, such
as the effects of the rework on the project schedule or on other project
activities, were not generally available and are not included in the deviation
costs presented herein. In addition, no data were available on the costs
associated with quality management activities.

TABLE 1. Descriptions of Projects Studied

Type of Type of Total installed

Project construction contract3 project cost
d) (2) (3) (4)
A Retrofit FP $ 20,000,000
B Retrofit FP $ 6,800,000
C New FP $ 25,000,000
D Upgrade FP $ 11,000,000
E Upgrade CR $100,000,000
F Upgrade CR $900,000,000
G New CR $100,000,000
H Retrofit CR 700,000 man-hours
I New FP $106,000,000
a
FP = fixed price; CR = cost reimbursement.

36

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 CLASSIFYING DEVIATION DATA

The deviation data collected on the nine projects were classified to allow
a more descriptive analysis of the causes of the deviations and their asso-
ciated costs. The deviation cost data were divided into five major areas—
design, construction, fabrication, transportation, and operability. Each of
these areas were further subdivided by type of deviation, i.e., change, error,
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

or omission, in accordance with the definitions in the QMTF glossary in

Appendix I.
For the projects studied, a large percentage of the deviations were due
to design changes (over 50% of the total number of deviations). To better
define the costs associated with design changes, the deviation category for
design changes was expanded to seven design-change categories.
The deviation categories are shown in Table 2 and discussed herein. The
categories are mutually exclusive, and each deviation should be classified
by only one of the categories.

Design Deviations
Design deviations are related to the design of the project. Design errors
(DE) are the result of mistakes or errors made in the project design. Design
omissions (DO) result when a necessary item or component is omitted from
the design. Design changes occur when changes are made in the project
design or requirements, and are further classified as follows.

Design Change/Improvement (DCI)

DCI includes only design revisions, modifications, and improvements
initiated through the design process. Examples of this are changes that are
the result of design reviews, model reviews, and technological advances.

TABLE 2. Deviation Categories Used for Classifying Project Data

Deviation category Brief description •

0)
Construction change
Construction error
Construction omission
Design change/improvement
Design change/construction
Design change/field
Design change/owner
Design change/process
Design change/fabrication
Design change/unknown
Design error
Design omission
Operability change
Fabrication change
Fabrication error
Fabrication omission
Transportation change
Transporation error
Transporation omission
(2)
Change in the method of construction
Error made during construction
Omission made during construction
Design revision, modifications, and improvements
Design change initiated by construction
Design change required due to field conditions (e.g.,
lack of as-builts)
Design change initiated by the owner
Design change initiated by operations or process
Design change initiated by the fabricator
Design change with an unknown source of initiation
Error made during design
Omission made during design
Change made to improve operability
Change made during fabrication
Error made during fabrication
Omission made during fabrication
Change made to method of transportation
Error made in method of transportation
Omission made in transportation

37

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 Changes initiated for any other reason are classified under one of the other
design change designations.

Design Change/Construction (DCC)

DCC is changes in design made at the request of the field or construction
personnel. An example of this is addition of concrete pads to permit proper
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

installation of equipment.

Design Change/Field (DCF)

DCF is design changes due to field conditions in retrofit and upgrade
projects. An example of this is when the existing structure, equipment, or
pipe location differs from the details given on available drawings, and the
deviation could not have been foreseen by the designer.

Design Change/Owner (DCO)

DCO is changes in the project design initiated by the owner. Examples
of this are a change in project scope or additional electrical outlets in an
office.

Design Change/Process (DCF)

DCP is design changes in the process portion of the facility initiated by
an owner's representative or consultant familiar with the expected opera-
tions and processes to be fulfilled by the facility. An example of this is the
addition of valves, pumps, electrical equipment, or instrumentation that
affect the operation of the completed facility.

Design Change/Fabrication (DCR)

DCR is changes in design initiated or requested by the fabricator or
supplier. An example of this is a fabricator request for a change in vessel
dimensions to provide uniformity between parts.

Design Change/Unknown (DCU)

DCU is design changes for which the description does not yield enough
information regarding the reason or source of change, and discussion with
the project representative affords no insight. An example of this is a change
with a description such as "structural steel design change." While this change
may have been an improvement in design or the result of a model review,
it may also have been a redesign due to an error.

Construction Deviations
Construction deviations are related to the construction phase of the proj-
ect and consist of those activities and tasks that take place at the project
site. A construction change (CC) is defined as a change in the method of
construction, such as placing concrete by pump father than by bucket.
Construction changes are usually made to enhance the constructability of
the project. Deviations classified as construction errors (CE) are the result
of erroneous construction methods or procedures. Construction omissions
(CO) are those deviations that occur due to the omission of some construc-
tion activity or task.

Fabrication Deviations
Fabrication deviations are related to shop fabrication. Changes, errors,
and omissions that occur during field fabrication are included in the con-
38

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 struction deviation categories. Fabrication changes, errors, and omissions
are those deviations that occur at, or are the result of, work performed by
a vendor, fabricator, or supplier. A change made in or during fabrication
is classified as a fabrication change (FC). Fabricated parts that are not in
accordance with the specifications are noted as fabrication errors (FE), while
parts or pieces that are included in the specifications but are not supplied
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

are denoted as fabrication omissions (FO).

Transportation Deviations
Transportation deviations are related to the transport of equipment, ma-
terials, or supplies. A transportation change (TC) indicates a change in the
method of shipment, e.g., shipping by air to expedite delivery rather than
shipping by truck. Transportation errors (TE) denote errors made in trans-
porting a product, e.g., shipping an article in separate pieces when the
specifications require the shipment of an assembled product. Transportation
omissions (TO) occur when a required part or item is not included in the
appropriate shipment.

Operability Deviations
A differentiation was made between changes, errors, and omissions made
to the operation or process portion of the facility and those changes made
to improve operability. An operations change might be the use of two pumps
instead of one, or the addition of check valves in a required line; while an
operability improvement might be relocating valve handles to improve op-
erator access. Changes in operability are denoted with the deviation code
(OC), while changes made in the operation or process portion of the facility
are included in a specific design-change category. There is no need for error
or omission categories for operability since errors and omissions in opera-
bility are the result of an error or omission made in design, fabrication, or
construction.

Decision Rules
If comparisons are to be made among different projects, it is essential
that a clear and consistent set of criteria or decision rules be established to
govern the way in which the deviation costs are assigned to the various
deviation categories. Appendix II presents the decision rules that were
developed to ensure consistent classification of each deviation. The devia-
tion categories are designed to be mutually exclusive and only one complete
category should be applied to each deviation (a single deviation report may
contain several different deviations) (see Fig. 1).
The primary descriptor designates the major area, i.e., design (D), con-
struction (C), fabrication (F), transportation (T), or operability (O). The
second descriptor designates whether the deviation was a change (C), error
(E), or omission (O). The tertiary descriptor is only used to further subdivide
the design changes category (Table 2).
In some cases it may not be possible to discern a single classification from

Primary descriptor Tertiary descriptor

XXX
I
Secondary descriptor
FIG. 1. Explanation of Deviation Category Descriptors

39

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 the information given in a deviation report. A general description such as
"structural steel change" with no further clarification could not be accurately
classified by a complete deviation category. In recognition of this possibility,
the "unknown" deviation category descriptors are indicated as alternatives
in the list of decision rules. These descriptors must not be used as "catch-
all" categories for difficult cases, and should be used only when no other
classification is possible.
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

ANALYSIS OF DATA

The data were analyzed both in terms of numbers and costs of deviations.
Since the size of each of the projects (in total cost) varied, comparisons of
number of deviations and deviation costs were all performed on a percentage
basis to allow comparisons among the projects. The analyses consisted of
the number of deviations, deviation costs as a percentage of total project
deviation costs, and deviation costs as a percentage of total project cost.

Number of Deviations
Table 3 presents the number of deviations in the design, construction,
fabrication, transportation, and operability areas as a percentage of the total
number of deviations on the project. The greatest number of deviations
occurred in the design and construction areas. Design deviations accounted
for 67-90% of the total number of deviations on the projects, while con-
struction deviations ranged from 5-29% of the total number of deviations.
Table 4 shows the percentage of the total number of deviations for design
changes, errors, and omissions for each project. Design changes generally
resulted in the greatest number of design deviations. An exception is project
I, which reported no design changes. Discussions with project representa-
tives revealed that for this project, changes in design were not recorded
unless they were the result of an error or omission.
Design changes for improvement (DCI) caused, on the average, the great-
est number of deviations, 13.3%, followed by design changes in process
(DCP), 10.9%, and design changes initiated by the owner (DCO), 9.9%,
while 13.2% of the total number of deviations were classified as design
changes of unknown origin (DCU).

Cost of Deviations
Table 5 presents the deviation costs for each area as percentages of the
total deviation costs for each project. Fig. 2 presents the nine-project av-
erages for the same data. Since design deviations accounted for such a large

TABLE 3. Number of Deviations as Percentage of Total Number of Deviations for

Each Project

Project
Aver-
Area A B c D E F G H 1 age
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Design 66.5 89.6 74.7 88.6 73.8 84.8 69.5 75.1 80.8 78.1
Construction 25.5 8.4 15.4 5.7 19.5 4.9 28.8 12.0 19.2 15.5
Fabrication 6.0 1.0 8.5 2.3 3.7 0.3 1.4 12.4 0.0 4.0
Transportation 0.7 1.0 1.1 1.1 0.6 0.0 0.0 0.0 0.0 0.5
Operability 1.3 0.0 0.3 2.3 2.4 10.0 0.3 0.5 0.0 1.9

40

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 TABLE 4. Number of Design Deviations as Percentage of Total Number of Devia-
tions for Each Project
Project
Deviation
category A B C D E F G H 1 Average
(1) (2) (3) (4) (5) (6) (7) (8) 0) (10) (11)
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

Design change/ 10.4 7.3 9.5 26.1 26.4 24.1 8.3 7.3 0.0 13.3
improvement
Design change/ 1.0 0.0 2.8 2.3 3.3 0.5 1.4 6.8 0.0 2.0
construction
Design change/ 14.8 11.5 0.0 1.1 2.7 1.9 0.0 4.7 0.0 4.1
field
Design change/ 10.4 18.8 9.7 9.1 14.6 5.7 0.2 12.1 0.0 9.0
owner
Design change/ 2.0 0.0 3.6 27.3 20.1 37.0 0.7 7.4 0.0 10.9
process
Design change/ 0.0 0.0 0.0 0.0 0.3 0.0 0.0 0.0 0.0 0.0
fabrication
Design change/ 2.7 9.4 11.5 17.1 0.3 9.7 54.1 14.2 0.0 13.2
unknown
Design change 41.3 46.9 37.1 83.0 67.8 78.9 64.7 52.4 0.0 52.5
total
Design error 24.8 26.0 24.3 5.7 4.0 5.7 4.7 21.6 60.1 19.7
Design omission 0.3 16.7 13.3 0.0 2.1 0.3 0.1 1.2 20.7 6.1
Design total 66.4 89.6 74.7 88.7 73.9 84.9 69.5 75.2 80.8 78.3

TABLE 5. Deviation Costs as Percentage of Total Project Deviation Costs

Project Aver-
Area A B C D E F G H 1 age
0) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Design 58.2 84.0 85.2 94.6 51.4 94.0 79.5 68.5 96.5 79.1
Construction 30.2 15.8 9.8 1.7 46.2 4.5 20.3 20.3 3.6 16.9
Fabrication 9.9 0.1 4.0 3.1 1.4 0.1 0.2 11.1 0.0 3.3
Transportation 0.8 0.1 0.8 0.0 0.1 0.0 0.0 0.0 0.0 0.2
Operability 0.8 0.0 0.2 0.6 1.0 1.5 0.0 0.1 0.0 0.5

percentage of the deviations, Fig. 3 presents a breakdown of the design

deviations. Deviation costs for the design-change categories amounted to
an average of 54.2% of the total deviation costs.
Table 6 presents the deviation costs for each project as a percentage of
total project cost. Fig. 4 shows the nine-project average deviation costs as
a percentage of total project cost. Total project deviation costs ranged from
0.4% to 25.9% of the total project cost, with an average of 12.4%. Design
deviations generally accounted for the greatest increase in total project cost,
ranging from 0.4% to 20.6%, with an average of 9.5%. The average values
for the design deviation categories are broken down in Fig. 5.
Construction deviations, averaging 2.5%, contributed less than design
deviations to the total project cost. However, the percentage of construction
deviations was generally larger than the combined percentages attributable
to fabrication, transportation, and operability deviations. Deviations in de-
41

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 Transp & Operability
0.7%
Construction
16.9%
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

Fabrication
3.3%

Design
79.1%

FIG. 2. Nine-Project Average for Deviation Costs as Percentage of Total Project

Deviation Costs

DCP

DCI
_ . --. DCU 14.3%
Design '"--..17.3%
Deviations
Design Changes
(54.2%)

See Table 2 for descriptions

of design changes.

FIG. 3. Nine-Project Average Design Deviations as Percentage of Total Project

Deviation Costs

TABLE 6. Deviation Costs as Percentage of Total Project Costs

Project
Aver-
Area A B C D E F G H 1 age
0) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11)
Design 4.2 12.1 4.4 20.0 11.8 8.8 20.6 3.5 0.4 9.5
Construction 2.2 2.3 0.5 0.4 10.6 0.4 5.3 1.0 0.0 2.5
Fabrication 0.7 0.0 0.2 0.7 0.3 0.0 0.1 0.6 0.0 0.3
Transoortation 0.1 0.0 0.1 0.0 0.0 0.0 0.0 0.0 0.0 0.0
Operability 0.1 0.0 0.0 0.1 0.2 0.1 0.0 0.0 0.0 0.1
Total 7.3 14.4 5.2 21.2 22.9 9.3 26.0 5.1 0.4 12.4

42

J. Constr. Eng. Manage., 1992, 118(1): 34-49

Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

Construction
2.5%

Other includes fabrication,

transportation & operability.

FIG. 4. Nine-Project Average Deviation Costs as Percentage of Total Project Cost

DCP DCF o.2%

v DCO 0.9%

Design
Deviations
Design Changes
(7.9%)

See Table 2 for descriptions

of design changes.

FIG. 5. Nine-Project Average Design Deviations as Percentage of Total Project

Cost

sign and construction, therefore, were the primary causes of both the number
of deviations and increases in project costs.

EFFECT OF TYPE OF CONSTRUCTION

Table 7 and Fig. 6 present a comparison of the average percentages of

the total number of deviations for the three types of construction; new,
retrofit and upgrade. Upgrade construction projects resulted in a higher
average percentage of the total number of deviations in the design change/
improvement (DCI) category than did new or retrofit construction. Upgrade
projects also exceeded new and retrofit projects in the average percentage
of the total number of design change/process (DCP) deviations. The higher
averages are not surprising since upgrade projects concern the process and
improvement of operating facilities.
Table 8 and Fig. 7 display the average deviation costs as percentages of
total project cost for each type of construction. The average deviation costs
43

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 TABLE 7. Average Number of Deviations as Percentage of Total Number of De-
viations for Each Project
Type of Construction
Deviation category New Retrofit Upgrade
(1) (2) (3) (4)
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

Construction change 13.7 13.6 8.9

Construction error 7.3 1.7 1.1
Construction omission 0.2 0.0 0.0
Design change/improvement 5.9 8.3 25.5
Design change/construction 1.4 2.6 2.1
Design change/field 0.0 10.3 1.9
Design change/owner 3.3 13.7 9.8
Design change/process 1.4 3.1 28.1
Design change/fabrication 0.0 0.0 0.1
Design change/unknown 21.9 8.8 9.0
Design error 29.7 24.2 5.1
Design omission 11.4 6.1 0.8
Other 3.8 7.6 7.6

• i Other
I I Construction
O 41 H H Design
CD 31
A
E 21
3
2
1(

New Retro-Fit Upgrade

Type of Construction
FIG. 6. Nine-Project Average Number of Deviations as Percentage of Total Num-
ber of Deviations

for design (Table 8) are consistent with the average number of deviations
for design (Table 7). Upgrade projects had the highest percentage of total
deviation cost in the DCI and DCP categories. Similarly, the retrofit average
percentage of total deviation cost in the DCF category exceeds the upgrade
and new construction averages.
Analysis of the construction deviations for the three types of construction
reveals that the new construction projects had the greatest average number
of construction errors and omissions, but that construction errors and omis-
sions on the retrofit projects had the greatest average impact on total de-
viation costs. Construction changes for the upgrade projects were not as
numerous as for the new and retrofit projects, but they had a greater impact
on total project costs for the upgrade projects. Thus, a better understanding
44

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 TABLE 8. Average Deviation Costs as Percentage of Total Project Cost

Type of Construction
Deviation category New Retrofit Upgrade
0) (2) (3) (4)
Construction change 1.8 1.7 3.7
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

Construction error 0.1 0.1 0.1

Construction omission 0.0 0.0 0.0
Design change/improvement 0.4 0.5 5.8
Design change/construction 0.1 0.1 0.2
Design change/field 0.0 0.5 0.1
Design change/owner 0.1 0.9 1.5
Design change/process 0.0 0.2 4.4
Design change/fabrication 0.0 0.0 0.0
Design change/unknown 6.7 1.5 0.8
Design error 1.0 2.4 0.6
Design omission 0.2 0.6 0.1
Other 0.1 0.4 0.5
Total 10.5 8.9 17.8

Total
Design
Construction
Other

New Retro-Fit Upgrade

Type of Construction
FIG. 7. Nine-Project Average Deviation Costs as Percentage of Total Project Cost

of possible problem areas can be gained through consideration of the dif-

ferences in the types of construction projects than can be gained from a
general summary of all project data.

ANALYSIS OF RESULTS

For the projects studied, identified deviation costs averaged 12.4% of the
total installed project cost. This value is only the tip of the iceberg. Dis-
cussions with representatives of the projects studied indicated that not all
of the deviations on a project were recorded. Construction changes made
at the site were often not included in formal reports, and errors were often
repaired and/or accepted immediately rather than expending the time and
effort to file formal requests. The deviation data gathered included only the
45

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 direct cost of the rework for the item in question and included no indication
of impact on the rest of the project. Therefore, both the number and costs
of deviations reported for the projects in this study are conservative esti-
mates of the actual values. The statistics reported in this paper confirm the
importance of identifying the causes of deviation costs so that they can be
reduced.
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

CONCLUSIONS

The two major areas resulting in deviations were design and construction.
Design deviations averaged 78% of the total number of deviations, 79% of
the total deviation costs, and 9.5% of the total project cost for the nine
projects studied. Furthermore, design changes accounted for two-thirds of
the design deviations. Construction deviations accounted for 16% of the
total number of deviations, 17% of the total deviation costs, and 2.5% of
the total project cost. These results indicate that rework costs are a signif-
icant portion of total costs.
The analyses presented in this paper demonstrate that a methodology to
categorize historical data can be used to identify the type and number of
quality deviations and their associated rework costs. The methods presented
herein cannot only identify deviation costs, but can also identify the most
likely areas in which to reduce them on various types of future projects.

ACKNOWLEDGMENTS

The writers gratefully acknowledge the Construction Industry Institute

(CII) for providing funding to support this research effort, and also extend
their sincere gratitude to the members of the CII quality management task
force for their guidance, support, and contributions. Additionally, thanks
are offered to those companies and individuals who contributed their rec-
ords, time, and experience.

APPENDIX I. GLOSSARY OF QUALITY TERMINOLOGY

• Acceptance criteria: specified limits placed on characteristics of a prod-

uct, process, or service defined by codes, standards, or other require-
ment documents.
• Appraisal: quality activities employed to determine whether a product,
process, or service conforms to established requirements, including de-
sign review, specification review, documentation review, constructa-
bility review, materials inspection/tests, craftsman testing, quality status
documentation, and postproject reviews.
• Audit: a formal, independent examination with intent to verify con-
formance with established requirements. An audit does not include
surveillance or inspection for the purpose of process control or product
acceptance.
• Change: a directed action altering the currently established require-
ments. Changes may encompass design, fabrication, construction, etc.
and materially affect the approved requirements, the basis of design,
the existing scope of the contract plans and specifications, or operating
capability of the facility.
• Corrective action: measures taken to rectify conditions adverse to qual-
46

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 ity, and, where necessary, to preclude repetition. Corrective action
includes rework for nonconformance and deviations.
9
Cost of quality: The cost associated with quality management activities
(prevention and appraisal) plus the cost associated with deviations.
<• Criticality: a measure of the significance or impact of failure of a product,
process, or service to meet established requirements.
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

s
Defect: a deviation of a severity sufficient to require corrective action.
e
Deviation: a departure from established requirements. A deviation may
be classified as an imperfection, nonconformance, or defect based on
its severity.
9
Deviation costs: the sum of those costs, including consequential costs
such as schedule impact, associated with the rejection or rework of a
product, process, or service due to a departure from established re-
quirements.
9
Error: any item or activity in a system that is performed incorrectly
resulting in a deviation, e.g., design error, fabrication error, construc-
tion error, etc. An error requires an evaluation to determine what
corrective action is necessary.
9
Imperfection: A deviation that does not affect the use or performance
of the product, process, or service. In practice, imperfections are de-
viations that are accepted as they are.
° Nonconformance: a deviation that occurs with a severity sufficient to
consider rejection of the product, process, or service. In some situations
the product, process, or service may be accepted as is; in other situations
it will require corrective action.
9
Omission: any part of a system, including design, construction, and
fabrication, that has been left out, resulting in a deviation. An omission
requires an appraisal to determine what corrective action is necessary.
9
Prevention: quality activities employed to avoid deviations, including
such activities as: quality systems development, quality program de-
velopment, feasibility studies, contractor/subcontractor evaluation, quality
orientation activities, and certification/qualification.
9
Project: all those elements associated with a facility from initial concept
to final disposition.
9
Project elements: the major phases of a project, which include: pre-
planning, design, procurement, construction, start-up, operation, and
final disposition.
9
Quality: conformance to established requirements (not a degree of
goodness).
9
Quality activities: those activities in a project directly associated with
prevention and appraisal.
9
Quality assurance: all those planned or systematic actions necessary to
provide adequate confidence that a product, process, or service will
conform to established requirements.
9
Quality control: inspection, test, evaluation, or other necessary action
to verify that a product, process, or service conforms to established
requirements.
9
Quality management: concerns the optimization of the quality activities
involved in producing a product, process, or service. As such, it includes
prevention and appraisal activities.
47

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 • Quality management costs: the sum of those costs associated with pre-
vention and appraisal activities.
• Quality performance management system: a management tool providing
data for the quantitative analysis of certain quality-related aspects of
projects by systematically collecting and classifying costs of quality.
• Requirement: a contractually established characteristic of a product,
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

process, or service. A characteristic is a physical or chemical property,

a dimension, a temperature, a pressure, or any other specification used
to define the nature of a product, process, or service.
APPENDIX II. DECISION RULES FOR CLASSIFYING DEVIATIONS BY
DEVIATION CATEGORY

To Determine the Primary Descriptor

1. Does the deviation report clearly indicate a single primary descriptor?
Yes: Select the indicated descriptor.
No: Proceed to step 2.
2. Is the description unclear or is more than one category indicated?
Yes: Proceed to step 3.
No: Select the closest acceptable descriptor.
3. Are separate costs indicated for each category?
Yes: Classify each of the separate costs with the indicated primary
descriptor.
No: Proceed to step 4.
4. Is a separate percentage of the single cost listed for each indicated item?
Yes: Classify each percentage item separately with the appropriate
primary descriptor.
No: Proceed to step 5.
5. Does the description indicate an overall similarity between the items?
Yes: Classify the entire cost according to the primary descriptor indi-
cated by the similarity.
No: Proceed to step 6.
6. Does the description of the deviation give a more elaborate discussion of
one item with the others treated secondarily?
Yes: Classify the entire cost by the primary descriptor indicated by the
detailed discussion.
No: Proceed to step 7.
7. Contact the project representative for help.
8. Based on the discussion with the project representative, is a primary de-
scriptor indicated?
Yes: Classify by the primary descriptor indicated by the project rep-
resentative.
No: Classify as unknown.
To Determine the Secondary Descriptor
1. Is the deviation clearly described by a single definition of change, error,
or omission?
48

J. Constr. Eng. Manage., 1992, 118(1): 34-49

 Yes: Classify by the appropriate secondary descriptor.
No: Proceed to step 2.
2. Does the description indicate multiple secondary descriptors?
Yes: Proceed to step 3.
No: Classify by the secondary descriptor that is most appropriate to
Downloaded from ascelibrary.org by Universitas Indonesia on 09/11/19. Copyright ASCE. For personal use only; all rights reserved.

the description.
3. Does the description of the deviation give a significant indication of one
secondary descriptor with the others treated in passing?
Yes: Classify the deviation with the secondary descriptor indicated by
the detailed discussion.
No: Proceed to step 4.
4. Contact the project representative for help.
5. Based on the discussion with the project representative, is a descriptor
indicated?
Yes: Classify by the secondary descriptor indicated by the project rep-
resentative.
No: Classify as unknown.

To Determine the Tertiary Descriptor (for Design Changes)

Classify each design change with the tertiary descriptor that most closely
describes the reason for the deviation.

APPENDIX III. REFERENCES

Burati, J. L. and Farrington, J. J. (1987). "Costs of quality deviations in design and

construction." CII Source Document No. 29, The Construction Industry Inst.,
Austin, Tex.
Davis, K., Ledbetter, W. B., and Burati, J. L. (1989). "Measuring design and
construction quality costs." /. Constr. Engrg. and Mgmt., ASCE, 115(3), 385-
400.
Diekmann, J. E., and Nelson, M. C. (1985). "Construction claims: Frequency and
severity." /. Constr. Engrg. Mgmt., ASCE, 111(1), 74-81.
Ledbetter, W. B. (1983). "A construction industry cost effectiveness project report."
Report A-6, Feb., Modern Management Systems, The Business Roundtable, New
York, N.Y., 1-39.
"Quality: The U.S. drives to catch up." Business Week, 100(18), Nov. 1, 66-80.
Shilstone, J. M. (1983). "Welcome, background and program objectives." Proc. Nat.
Conference on Quality Assurance in the Bldg. Community, Shilstone & Associates,
Inc., July, 1-6.

49

J. Constr. Eng. Manage., 1992, 118(1): 34-49