A METHOD FOR ANALYSING DELAY DURATION

CONSIDERING LOST PRODUCTIVITY THROUGH

CONSTRUCTION PRODUCTIVITY DATA MODEL
Han-Guk Ryu1, Hyun-Soo Lee2 and Jung-Ho Yu3
Department of Architecture of Seoul National University, 151-742, Shilim-dong 56-1, Kwanak-gu,
Seoul, Korea

A delay claim occurs due to the difference between the actual completion date and the
contract completion date. Though the responsibility and duration of a delay is very
important information on analysing a delay claim, it is generally difficult task because
there are many factors complicatedly affecting the delay. Lost productivity or loss of
productivity is one of major factors causing a delay. However, there are few studies
on converting lost productivity into delay duration. Thus, to calculate fairly the delay
duration of the liquidated damage, it is needed a calculating method related to many
impact factors and their impacted productivity. This study proposes a method for
analysing delay duration considering lost productivity by using strong points of
schedule such as CPM, Bar chart consistent with the project phases and construction
productivity database maintained data consistency and minimized data redundancy.
The main outcomes of this study are: (1) a method of calculating delay duration
caused by lost productivity, (2) a data model that is a basis of the database containing
information regarding lost productivity. With these outcomes, delay durations and
responsibilities could be more accurately calculated.

Keywords: claim, database, data model, delay, lost productivity.

INTRODUCTION
In construction contracts, like quality and cost of construction, construction duration is
the basic component and the item, which should be managed importantly. If a
contractor can work in the condition supposed to complete the construction project
before constructing the structure, a contractor can finish timely the construction
project. However, it is harder than any other industries to complete the construction
project in which many construction trades participated during construction phase. In
most cases, there are lots of unknown variables that occur during the construction
process and hinder the timely completion of a project.
Lost productivity is also a major cause of a delay. However, there have been few
studies on converting the lost productivity into the work delay duration. The schedule
delay calculation related to the various, serial, complex delay impact factors and lost
productivity must be a complicated problem hard to solve and prove. The established
tools for calculating delay claims are usually CPM networks or Bar charts. But the
tools have a limitation to calculate exactly. So Knoke and Jentzen argue that the
computer-based As-Built Schedule Database is more useful tool for the investigation
1
hglyu22@snu.ac.kr
2
hslee@gong.snu.ac.kr
3
azure88@snu.ac.kr
Ryu, H-G, Lee, H-S and Yu, J-H (2003) A method for analysing delay duration considering lost
productivity through construction productivity data model. In: Greenwood, D J (Ed.), 19th Annual
ARCOM Conference, 3-5 September 2003, University of Brighton. Association of Researchers in
Construction Management, Vol. 2, 735-43.
Ryu, Lee and Yu

of project performance (1994). The facts recorded in the database may be used to
identify actual sequential relationships between project activities and the time frames
during which the activities were performed and analysing delay claims, productivity,
scope change, payment issues, etc.

REVIEW OF PREVIOUS SCHEDULE DELAY CALCULATING

METHODS
Many studies or methodologies directed at analysing delay and lost productivity have
been reported (Kallo 1996; Bubshait and Cunningham 1998; Al-Saggaf 1998; Kartam
1999; Finke 1999; Reichard and Norwood 2001).The problems of delay calculating
studies related to lost productivity can be summarized as three cases as follows. (1)
Established calculating studies of delay related to lost productivity, which are limited
to studies of converting lost productivity into cost using such as measured mile
analysis and the productivity data published from CII, NECA and MCAA. (2) Delay
causes are conceived as activity in a project schedule such as a method of What-If
evaluation, But-For schedules, affected baseline schedule and collapsed as-built
analysis. (3) The impacted activities are analysed in the form of activities that have
non-impacted productivity. However, if some variables or impact factors impact the
next work in the sequence, the impacted work must be lost productivity work. As a
result, a study concerning methods of calculating schedule delay considering lost
productivity is not sufficient.

BASIC CONCEPTS
Productivity may be defined as the quantity of work produced per man-hour,
equipment hour, or crew hour (Finke 1998).As shown in Figure 1, it can be said that
the lost productivity is the productivity impacted adversely by unexpected factors or
impact factors. For example, a curtain wall crew consisted of 5 workers installing
34.65m2 per hour can be said to have a productivity rate of 6.93m2xman/hour under
good conditions not influenced by any other impact factor. But if a work affected by
any impact factor such as unexpected adverse weather, it will take some times or days
for the impacted work productivity to be the un-impacted work productivity or the
planned work productivity. The work productivity will be declined.
To calculate fairly the delay of the liquidated damage, it is needed a calculating
method related to many impact factors and their impacted productivity. The following
concepts are employed:
1. Planned Work Duration (PWD) is the work duration with the planned productivity.
2. Actual Work Duration (AWD) is the work duration with the actually un-impacted
productivity obtained from the entire period of work duration.
3. Start Time Variance (STV) is the difference between the actual start time of a work
and the finish time of the preceding work on an as-built schedule
4. Finish Time Variance (FTV) is the difference between the contractor's AWD and
PWD.
5. Lost Productivity Quantity (LPQ) denotes the work quantity, which could be
finished during un-impacted work duration.
6. Lost Productivity Duration (LPD) can be defined as opportunity duration could be
worked as much as LPQ.

736
 Delay duration

FTV
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41
Fl.

Moved average productivity Average productivity

of as- planned schedule of as- built schedule

Average productivity
of as- planned schedule Lost productivity
(Loss of productivity)

nonwork day productivity

24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

STV day
Figure 1: Lost productivity

PROCESS FOR DELAY DURATION ANALYSIS

A common method of calculating delay duration is performed by comparing as-
planned schedule and as-built schedule written by CPM (Kraiem and Diekmann 1987;
Bubshait and Cunningham 1998). Critical path is changed not commonly by schedule
delay and acceleration (Arditi and Robinson 1995). The works on the critical path
influence ultimately the delay of project completion.
Construction works consists of millions or thousands of works and the works might be
performed usually in different conditions from the point of planning. Those works can
be classified by two types. The one type is impacting works caused adversely by
impact factors to the completion day. The other type is un-impacting works to the
completion day. So it is needed to analyse which work impact the completion date of
the project and its impact degree. And then we can examine the responsibility of the
delay causes.
An analysis of the delay duration could be processed according to the following
procedure. The method process of calculating delay duration considering lost
productivity is shown in Figure 2.

EQUATIONS FOR CALCULATING WORK DELAY DURATION

Lost Productivity (LP) means the loss of productivity caused by unknown variable or
Impact Factor (IF) occurring during the construction phase. Lost Productivity
Duration (LPD) can be defined as opportunity duration as much as Lost Productivity
Quantities (LPQ), which could be worked during un-impacted, works duration.

737
Ryu, Lee and Yu

Pro p rie ty a n a lysis o f

a s- p la n n ed sc hed u le
a n d a s- b u ilt sc h ed u le

C ritic a l wo rk a n a lysis
o f a s- b u ilt sc h ed u le

C a lc u la tio n o f sta rt tim e va ria n c e

a n d fin ish tim e va ria n c e

C a u ses a n a lysis o f s ta rt tim e va ria n c e

a n d fin ish tim e va ria n c e

P la n n ed wo rk d u ra tio n =
Ac tu a l wo rk d u ra tio n
- No n wo rk d a y d u ra tio n

YES
NO

Fin d in g th e evid en c e
o f fin ish tim e va ria n c e
NO
YES
C a u ses a n a lysis o f wo rk d ela y

C h a ra c teristic a n a lysis
o f im p a c t fa c to rs

C a lc u la tio n o f re a l lo st p ro d u c tivity
a n d su m o f rea l lo st p ro d u c tivities

S u m a ssig n m en t
o f th e rea l lo s t p ro d u c tivity d u ra tio n

C o m p le tio n The o wn ers resp o n sib ility Th e 3rd resp o n sib ility Th e c o n tra c to rs resp o n sib ility

Figure 2: The calculation process of delay duration

Table 1: The variables used in equations

Work Status Work Work Daily Work
Quantity Productivity Average Duration
Labours
Planned work Q0 P0 L0 D0
Work of unimpacted- duration Qu Pu Lu Du
Work of impacted duration Qi Pi Li Di

The LP of some activity can be calculated like Eq. (1) by the difference between the
un-impacted productivity of the activity and the impacted productivity of the activity.
The LPQ can be calculated like Eq.(2) by LP multiplied by Li(worked labours) during
Di(impacted work duration). The LPD can be calculated like Eq.(3) by LPQ divided
by the product of daily average labours during impacted work duration and un-
impacted productivity. The variables are in Table 1.

LP = (P P ) ............................................................................................ (1)
u i

LPQ = (P P ) L D .............................................................................. (2)

u i i i

738
 Delay duration

LPQ (P P ) L D ................................................................... (3)

LPD = = u i i i

Li P u LP i u

Could Be Duration(CBD) denotes the duration within which a work could be finished
with the daily average labours and the un-impacted productivity. This can be
calculated by Eq.(4), where Q0 denotes the planned work quantity, Qu denotes the
quantity worked in the normal and realistic work conditions of an un-impacted work
duration, and Qi denotes the quantities worked in the impacted work duration.

=
Q 0
=
Q + Q ................................................................................ (4)
u i
CBD
LP
u u LP u u

Contractors Duration Difference (CDD) compared with the P0 and the Pu denotes the
difference CBD and D0 by Eq.(5), where P0 denotes the planned work productivity
and D0 denotes the planned work duration.
Q .................................................................. (5)
CDD = CBD D 0
= D 0
0
LP u u

Work Delay (WD) consists of CDD, LPD and as shown in Eq.(6), where CDD and
LPD are independent variables and is an extraneous variable that accounts for any
delays other than CDD and LPD.

WD = CDD + LPD + ............................................................................. (6)

CAUSE ANALYSIS OF WORK DELAY DURATION

After the FTV evidences are confirmed, the impact factors can be analysed by
comparing the AWD with the PWD. WDs are classified into three cases: Including all
contractors cause and LPD; Including only LPD; Including LPD less CDD.

Including all contractor's cause and LPD

WD includes all delay caused by contractors fault and LPD during impacted work
duration as shown in Figure 3.
Start Point Planned Finish Point Actual Finish Point

Planned Work Duration WD(Work Delay)

As-Built Work Duration

CDD
LPD

CBD

: CDD(Contractors Duration Difference)

: LPD(Lost Productivity Duration)
: (Extraneous Variable)
: Impacted time of impact factor

Figure 3: Including all contractors cause and LPD

When CDD is larger than 0, as shown in Eq.7, the delay is caused by the contractors
mistakes such as allocation of lower labours and productivity than as-planned
schedule. In that case the contractor could not be compensable and should compensate
the owner for liquidated damage.

CDD = CBD D 0 ................................................................................ (7)

0

739
Ryu, Lee and Yu

Including only LPD

The work delay includes only LPD in the case that the actual work duration of the as-
built schedule is equal to the work duration of the as-planned schedule as shown in
Figure 4.
Start Point Planned Finish Point Actual Finish Point

Planned Work Duration WD(Work Delay)

As-Built Work Duration

LPD

CBD

CDD = 0

Figure 4: Including only LPD

When CDD is 0, as shown in Eq.8, the delay is caused by only lost productivity.
Because the planned work productivity is equal to the actual work productivity,
contractor doesnt have any responsibility.

CDD = CBD D = 0 .............................................................................. (8)

0

Including LPD less CDD

When CDD is smaller than 0, as shown in Eq.9, WD is calculated by only LPD. If it
were not for the LPD, the actual work could be finished earlier than planned work
duration, as shown in Figure 5.
Start Point Planned Finish Point Actual Finish Point

Planned Work Duration WD(Work Delay)

As-Built Work Duration

CDD
LPD

CBD

Figure 5: Including LPD less CDD

The CBD is shorter than the planned finish date. The work might be performed with
acceleration or more labours and productivity than planned. But the effort cannot be
recognized as a compensable delay.

CDD = CBD D 0 ................................................................................ (9)

0

CONSTRUCTION PRODUCTIVITY DATA MODEL

The proposed calculating method is based on the construction productivity database
which construction data can be cumulated in during the construction phase. Generally,
the evident claim data such as contract documents, drawings, specifications, and any
other documents related to contracts can be used to prove the direct impact causes, but
they are not sufficient to prove the indirect impact causes such as lost productivity.
The proposed calculating method requires a construction productivity database
containing related data to be collected during a construction phase. The below are a
conceptual and a logical construction productivity data model that were developed to
be used conceptually in this study.

740
 Delay duration

Previous study of construction claim database

The method for calculating schedule delay is based on the data. Database can be used
to collect diligently the data, which cannot be cumulated during construction phase.
However schedule delay studies using database are insufficient. Knoke and Jentzen
(1994) pointed out the limitation of the CPM network schedule as delay calculating
tool and emphasized the reliability of evidence maintained in the database during the
construction phase.

Construction productivity data model

A typical implementation of the data model includes a collection and analysis of data
requirements, conceptual data modelling and logical data modelling. Logical data
modelling can be performed based on the real world and change a substance into a
structured and detail information. Physical data modelling can be performed based on
the system.

Collection and analysis of data requirements

The most important thing of data model is collection and analysis of the requirements.
So the requirements are collected through the existing document, questionnaire and
interviews with experts. The requirements can be summarized the entity or the attrib-
tes of the entity: space, work, work day, method, subcontractor, drawing number,
spec, check list, material, documents, labour, machine, impact factor and so forth.

Conceptual construction productivity data model

Conceptual data model is based on examining the relations of entities for making an
ER Model through analysing the requirements. Conceptual data model for measuring
the construction productivity can be drawn as Impact Factors (IF) impacts the
Work_Space as a relation of Work and Space like Figure 6. The works are associated
with the Spec., Drawing, Machine, Material, Checklist and performed according to the
space order. Work_Space has documents related to as participants change orders and
subcontractors labours, labour information, IF information, space, etc.

Impact IF code #
Factor
Subcon code #

Checklist code #
Material code #
Subcon Labor code # IF_W_S IF_W_S code #

Checklist
Material
Labor
Work Day
Work code #
Work Work_Space
Spce.
Work_Space code #
Space
Spec. code #
Drawing Machine Document
Space seq. code #
Drawing code # Machine code # Document code #

Figure 6: Conceptual construction productivity data model

Logical construction productivity data model

Logical data model can be defined as a describing process for ER Model of conceptual
data modelling output to match systems such as DBMS (Data Base Management
System). Logical construction productivity data model can be designed like Figure 7

741
Ryu, Lee and Yu

through the definition of entities, attributes, identifiers and relations, normalization

and generalization.
Bui l di ng Act i vi t y
Pr oj ect _wor k
Act _code I F_appl i cat i on
Bl dg code Pr j _wor k_code Pr j _wor k_code ( FK) I F_appl _code
Bl dg_l ar ge_name Wor k_cl f _code ( FK) Pr j _Bl dg_name_code ( FK)
Pr j _Bl dg_name_code ( FK) Act _assi gnment _code ( FK)
Bl dg_medi um_nam Bl dg_code ( FK) Space_cl f _code ( FK)
Bl dg_smal l _name Bl dg_code ( FK) Pr j _code ( FK)
Pr j code ( FK) Act _code ( FK)
Wor k cl f code ( FK) Pr j _wor k_code ( FK)
Act _name Pr j _Bl dg_name_code ( FK)
Pl anned_dur at i on Bl dg_code ( FK)
Act ual _dur at i on Pr j _code ( FK)
Pl anned_st ar t _dat e Wor k_cl f _code ( FK)
Pl anned_f i ni sh_dat e Labor _code ( FK)
Act ual _st ar t _dat e Subcon_code ( FK)
Wor k_cl assi f i cat i on Act ual _f i ni sh_dat e I F code ( FK)
Wor k cl f code Pl anned_cost
Act ual _cost
Wor k_l ar ge_cl f
Wor k_medi um_cl f
Wor k_smal l _cl f
Subcon_code
Pr oj ect _Bui l di ng_name Act i vi t y_assi gnment
Pr j _Bl dg_name_code Act _assi gnment _code I mpact _Fact or
Bl dg_code ( FK) Space_cl f _code ( FK)
Act _code ( FK) I F code
Pr j code ( FK)
Pr j _wor k_code ( FK) I F_name
Labor Pr j _Bl dg_name_code ( FK) Regi st er ed_dat e
Labor _code Bl dg_code ( FK) I F_st ar t _dat e
Wor k_cl f _code ( FK) Pr j _code ( FK) I F_f i ni sh_dat e
Subcon code ( FK) Wor k_cl f _code ( FK) I F_dur at i on
Labor _name Labor _code ( FK)
Ski l l ed_l abor _bei ng Subcon code ( FK)
SSN A_D_Q_code
Bl ood_t ype
Pr oj ect Saf et y_educat i on_exi st ence
Pr j code
Pr j _name
Pr j _si t e
Gr ound_scal e Owner _r esponsi bi l i t y cont r act or _r esponsi bi l i t y Thi r d_r esponsi bi l i t y
Act i vi t y_dat e_Quant i t y
Under gr ound_scal e I F code ( FK) I mpact f act or code ( FK) I F code ( FK)
Tot al _Bl dg_ar ea Act _D_Q_code
Subcont r act or Space_Cl assi f i cat i onAct _assi gnment _code ( FK) Owner _r esp_name Cont r act or _r esp_name Thi r d_r esp_name
Si t e_ar ea
Cl i ent Subcon code Space cl f code Act _code ( FK)
Cont r act or Space_l ar ge_name Pr j _wor k_code ( FK)
Subcon_name Pr j _Bl dg_name_code ( FK)
Pl anned_st ar t _t i me Space_medi um_name
Pl anned_f i ni sh_t i m Space_smal l _name Bl dg_code ( FK)
Act ual _st ar t _t i me Pr j _code ( FK)
Act ual _f i ni sh_t i me Wor k_cl f _code ( FK)
Space_cl f _code ( FK)
Labor _code ( FK) Subconst r act or _r esponsi bi l i t y
Subcon code ( FK) I mpact f act or code ( FK)
Act _dat e Subcon_r esp_name
Act _Q

Figure 7: Logical construction productivity model

CONCLUSIONS
The purpose of this study is a method for calculating delay duration considering lost
productivity by using strong points of schedule such as CPM, Bar Chart consistent
with the project phases and construction productivity database maintaining data
consistency and minimizing data redundancy. The main outcomes of this study are as
follows. (1) The analysing process and the calculating method was presented for
calculating delay duration considering lost productivity. (2) Data model was
developed for recording and managing the items hard to be expressed in detail
schedule on construction phase. The required information for construction
productivity database was analysed and then construction productivity data model was
developed through a conceptual and a logical data model by extracting the entities and
attributes from the required information.

ACKNOWLEDGEMENT
The writers thank the Korea Institute of Science and Technology Evaluation and
Planning (Grant M10104000274-01J000012100) for supporting portions of the work
presented here.

REFERENCES
Abdulaziz A. Bubshait, Member (1998) Comparison of Delay Analysis Methodologies,
Journal of Construction Engineering and Management, Vol.124, No.4, 315-322.
Abdul-Malak, M A U, El-Saadi, M M H and Abou-Zeid, M G (2002) Process model for
administrating construction claims, Journal of Management in Engineering, ASCE,
18(2), 84-94
Al-Saggaf, H A (1998) The five commandments of construction project delay analysis, Cost
engineering, 40(4), 37-41

742
 Delay duration

Arditi, D and Robinson, M A (1995) Concurrent delays in construction litigation, Cost

engineering, 37(7), 20-30
Bubshait, A A and Cunningham, M J (1998) Comparison of delay analysis methodologies,
Journal of Construction Engineering and Management, ASCE, 124(4), 315-322
Cox, R K (1997) Managing change orders and claims, Journal of Management in
Engineering, ASCE, 13(1), 24-29
Finke, M R (1998) A better way to estimate and mitigate disruption, Journal of Construction
Engineering and Management, ASCE, 124(6), 490-497.
Kallo, G G (1996) Estimaing loss of productivity claims, Journal of Management in
Engineering, ASCE, 12(6), 13-15
Kangari, R (1995) Construction documentation in arbitration, Journal of Construction
Engineering and Management, ASCE, 121(2), 201-208.
Kartam, S (1999) Generic methodology for analysing delay claims, Journal of Construction
Engineering and Management, ASCE, 125(6), 409-419
Knoke, J R and Jentzen, G H (1994) Facilitate delay claim with as-built schedule databases,
Computing in civil engineering, Khozeimeh, K, ed., ASCE, Reston, Va., 2104-2111
Kraiem, Z M and Diekmann, J E (1987) Concurrent delays in construction projects, Journal of
Construction Engineering and Management, ASCE, 113(4), 591-602
Reichard, D D and Norwood C L (2001) Analysing the cumulative impact of changes, AACE
International Transactions
Shi, J J, Cheung, S O and Arditi, D (2001) Construction delay computation method, Journal
of Construction Engineering and Management, ASCE, 127(1), 60-65
Vidogah, W and Ndekugri, I. (1997) Improving Management of Claims, Journal of
Management in Engineering, Vol.13, No.5, 37-44.
Wickwire, J M and Smith, R F (1974) The use of critical path method techniques in contract
claims, Public Contract Law, 7(1), 1-45.

743