Journal of Financial Management of Property and Construction

A new framework for time-cost trade-off considering float loss impact

Sameh Monir El-Sayegh Rana Al-Haj
Article information:
To cite this document:
Sameh Monir El-Sayegh Rana Al-Haj , (2017)," A new framework for time-cost trade-off considering float loss impact ",
Journal of Financial Management of Property and Construction , Vol. 22 Iss 1 pp. -
Permanent link to this document:
http://dx.doi.org/10.1108/JFMPC-02-2016-0007
Downloaded on: 11 March 2017, At: 04:59 (PT)
References: this document contains references to 0 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 4 times since 2017*

Access to this document was granted through an Emerald subscription provided by emerald-srm:126209 []
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

 A new framework for time-cost trade-off considering float loss impact

Introduction

The construction industry witnessed an increase in the size and complexity of

construction projects during the past years. The increase necessitates the use of proper

and innovative project management techniques to handle the challenges and the risks

associated with projects. A project is defined as as a temporary endeavor undertaken to

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

create a unique product, service or result (PMBOK, 2013). Project management aims to

deliver the construction project on time, within budget while meeting or exceeding

customer expectations. One way to achieve the delivery of the project at the required

completion date and with the least cost is by the use of the least-cost scheduling

technique or the time-cost trade-off techniques.

Time-cost trade-offs are one of the most frequent and critical decisions that

project managers usually make. Project managers are frequently required to make time-

cost trade-offs (Liu and Rahbar, 2004). Time-cost trade-off refers to the method where

the project duration is shortened with a minimum added cost (Hegazy and Menesi,

2012; Chassiakos and Sakellaropoulos, 2005; Charoenngam and Popescu, 1995). In

general, project time and cost are linked via a relationship. As the project time is

shortened, the direct cost of critical activities increase, while the indirect cost

(overheads) of the project decreases. When there is a need for crashing the activities and

accelerating project completion, Larson and Gray (2014) mention several options based

on the resources constraints. Options when resources are not constrained include

outsourcing the work like by subcontracting, having overtimes or multiple shifts at

1
 work, or adding additional resources such as extra labors or extra machineries. On the

other hand, the available options when resources are constrained can include reducing

the overall scope of the project, or go with the fast tracking option by changing the

logical relationships between the activities in a way that the critical activities are

performed in a parallel rather than in a sequential basis.

Optimizing the project’s duration while maintaining the least crashing cost is

usually needed in order to complete the project activities earlier than originally
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

scheduled or to meet the project deadline with the least additional cost (Larson and

Gray, 2014). When the project duration is reduced, the total float available for

noncritical activities is reduced as well. Typically, the optimization process involves

deterministic procedure that is carried out until an optimum value is reached (Hinze,

2012). Available time-cost optimization methods are based on the concept of shortening

the duration of the critical activities in the network progressively while observing the

decrease in total project cost until the optimal solution that provides the shortest project

duration with the minimum total cost is reached (Hinze, 2012). The deterministic

optimization technique doesn’t consider the impact of the float loss within the

noncritical activities when the project duration is being crashed or reduced. Such losses

in total float can impact the project cost and schedule, and may lead to delays in

activities that are in a path; causing a ripple effect on the downstream activities of that

path; and therefore, losing the chance of early finish for these activities.

Construction projects are risky. Risk management is an essential and integral

part of project management on virtually all construction projects (Choudhry et al.,

2014). Project risk management is an important topic for practitioners and academic

2
 scholars (Zhang, 2016). Risk assessment is an integral part of project risk management

(El-Sayegh and Mansour, 2015). Risk quantification in construction schedules is usually

performed using Monte Carlo Simulation (MCS). When a project is scheduled, the

normal schedule will have certain risks that are measured by the probability of

completing the project on time. As the project duration is shortened, to reduce total cost,

the float is lost resulting in more critical or nearly critical activities. This, in turn, results

in reducing the probability of completing the project on time and increases the risk of
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

schedule delays. There is a need to incorporate that extra risk into the time-cost trade-

off problem. This will result in a more reliable project schedule. This paper proposes a

new framework that incorporates such impact into the time-cost trade-off process.

Time-cost trade-off techniques

There are several time-cost trade-off techniques that were developed over the past years.

The traditional time-cost trade-off technique is based on the Critical Path Method

(CPM). Time-cost optimization is based on the idea of shortening the critical activities

with the minimal increase in cost per unit of time (Charoenngam and Popescu, 1995).

Hinze (2012) explained a logical method for crashing that considers the schedule

network as a rigid frame. This technique is based on the use of link lag values that helps

in determining the possible number of times the activity can be crashed. Maximum

flow-Minimal cut is another manual optimization method for construction schedules

that was discussed by Liu & Rahbar (2004).

Several researchers used mathematical techniques to solve the time-cost

optimization problem. These methods include linear programming (Perera, 1980,

3
 Deckro et al., 1995, Islam et al., 2004; Chassiakos & Sakellaropoulos, 2005), Non-

linear programming (Klansek & Psunder, 2008), and mixed integer programming (Liu

et al., 1995; Moussourakis & Haksever 2007). Ammar (2011) developed a nonlinear

optimization model for project time cost tradeoff with discounted cash flow.

Optimization techniques require a lot of calculation time. This prompted researchers to

look for meta-heuristic techniques that can be used to determine the near optimum

solution in less time. Genetic Algorithms (GA) were used by Li & Love (1997), Leu &
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Yang (1999) and Que (2001). Fuzzy logic was used by Castro-Lacouture et al. (2009) to

incorporate the time-cost trade-off into the schedule assuming linear fuzzy relations.

Zahraie and Tavakolan (2009) used fuzzy genetic algorithms and fuzzy sets to solve the

time-cost-resource utilization optimization problem. Other meta-heuristic techniques

include the use of ant-colony optimization (Ng & Zhang, 2008; Afshar et al. 2009),

Tabu search (Hazir et al. 2011), particle swarm optimization (Zhang & Li, 2010 and

Yang, 2007), Harmony search (Geem, 2010) and simulated annealing (Sonmez and

Bettemir, 2012). Few authors considered risk in the time-cost trade-off problem. There

is a need to incorporate uncertainties in the time-cost trade-off problem (Eshtehardian et

al. 2008). Feng et al. (2000) developed a stochastic time-cost trade-off approach for

construction projects. Isidore & Back (2001) proposed an improved model in which

variability of activities in terms of cost and time is taken into consideration. Yang

(2005) proposed a chance-constrained programming model. Zheng and Ng (2005)

developed a stochastic time-cost optimization model using fuzzy sets. Al Haj and El-

Sayegh (2015) developed a new model for time-cost optimization considering float

consumption impact. However, in their model, they used a deterministic equation to

4
 calculate the float consumption impact. In this paper, the float loss cost is calculated

stochastically using Monte Carlo Simulation.

Float Loss Impact

The float is a measure of the schedule flexibility. Total float can be defined as the

amount of time an activity can be delayed without delaying the project. On the other
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

hand, the free float can be defined as the amount of time an activity can be delayed

without delaying the immediate successor activity (Charoenngam and Popescu, 1995).

Total float should be distinguished from free float as total float belongs to the activities’

path (it is shared by all the activities in the same path), while the free float belongs to a

particular activity. Since free float can’t be shared within the activities, its importance

compared to the total float is limited. Free float can be useful when selecting activities

for resource leveling. As float is a key element in project scheduling, several studies

were raised over this topic over the years. Ziegler (1985) introduced the minimal and

maximal float concept. The concept of minimal float represents the float present in

“worst case” while maximal float represents the float available in the “best case”

(Ziegler, 1985). Gong (1997) developed a method to find the optimum float use in a

project network. Optimum float was defined as the “point at which the sum of the cost

resulting from float use and the cost resulting from project delay due to float use is the

lowest (Gong, 1997). Sakka & El- Sayegh (2007) developed a method to control the

risks associated with the float loss due to delays in construction projects and its effect

on noncritical activities.

5
 Float ownership has been subject of debate over the last decades. Wickwire et

al. (1999) recognized that the float is an expiring resource that doesn’t belong to any

party but at the same time it is available to be used by the project parties on a fair basis.

Several approaches can be used in the construction industry to allocate float. Ponce de

Leon (1986) recommended that the total float of each activity is represented as a bar in

the bar chart to observe the critical and noncritical path delays. Householder and

Rutland (1990) developed the contract risk approach where the float is owned by the
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

party that assumes full responsibility for the project risk. De La Garza et al. (1991)

emphasized that the total float should be treated as a commodity that is traded between

the owner and the contractor. Gong and Rowings (1995) introduced the concept of safe

float to project scheduling. This method specifies the range of safe float to be used by

the project parties in general so as to logically minimize the risks associated with delays

in noncritical activities, based on a time-disturbance analysis over the project schedule

(Gong and Rowings, 1995). De la Garza et al. (2007) recommended that the total float

is distributed between the project parties based on a pre-agreed ratio to be stated in the

contract clauses. Al-Gahtani (2009) developed the total risk approach to assign the float

to parties based on the amount of risk they encounter in the project.

Al-Gahtani and Mohan (2005) emphasized that the float on noncritical activities

leads to efficient resource utilization. The amount of float is essential for resource

leveling and time-cost trade-off (Ammar, 2003). Gong (1997) emphasized that floats

are considered as safe harbor for resource allocation and other purposes without causing

a negative impact on the project duration. Al-Gahtani and Mohan (2005) stated that

each day of the total float available to a contractor is real money. De la Garza et al.

6
 (1991) recommended having a trade-in value for float while Gong (1997) recommended

limiting the use of float to reduce the risk of schedule overruns. Sakka and El-Sayegh

(2007) proposed a framework for quantifying the float loss impact on project time and

cost.

The float loss impact is not easily identified as a quantitative value unless the

schedule delay risk due to crashing is realized. During time-cost trade-off, critical

activities are crashed to reduce the project duration. As the project duration is reduced,
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

the float of non-critical activities is also reduced. The probability of completing the

project on time is also reduced. If we keep the probability constant, we get a new

duration. The difference between the new duration and the crashed duration represents

the impact of float loss. Then multiplying that difference by the indirect cost per day

will give us the float cost for each crash. For stochastic schedule analysis, the software

@Risk (an add-on to Microsoft Excel) is used to perform Monte Carlo Simulation

(MCS). To use the software, the project network needs to be modeled in Excel with

activity duration represented as Probability Density Function (PDF). In this paper, it is

assumed that all activities follow normal distribution functions with means and standard

deviations. However, the activity duration can be any distribution that the project

manager sees fit. Additionally, there is a need to specify the number of iterations. In this

research, we used 10,000 runs. However, the project manager may choose different

number of simulation runs as appropriate to the size of the project. For large projects,

the computational time will increase significantly by increasing the number of

simulation runs.

7
 The proposed method for calculating float cost can be explained using a simple

example. Assume that the normal deterministic project duration is 40 days. Considering

the risks, the mean duration and standard deviation is 41.39 days and 4.9 days,

respectively (after running MCS on the normal schedule). The probability of completing

the project on time (within 40 days) is calculated as 0.388 (assuming normal

distribution). If the project is shortened by one day, the new duration is 39 days.

However, the probability of completing the project on time (within 39 days) is reduced
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

to 0.349 (after running MCS on the crashed schedule). This reduction represents the

added risk due to crashing and the resulting float loss. If the project manager wants to

maintain the probability of project completion at 0.388, the corresponding duration is

39.52 days. Thus, the difference is 0.52 days. Assuming that the indirect cost is $2,000

per day, the resulting extra cost due to increased risk (float loss) is $1,045.

Proposed time-cost trade-off framework

As the project duration is shortened, the available float for non-critical activities is

reduced which increases the risk of not completing the project on time. A new

framework is proposed that enables calculating the effect of this float loss and includes

its effect in the trade-off process. Figure 1 illustrates the framework steps via a

flowchart. The calculations are performed in two stages: normal schedule analysis and

schedule compression analysis.

<Insert Figure 1 here>

Figure 1. Proposed time-cost trade-off framework

Stage 1- Normal schedule analysis

8
 In stage 1, the normal deterministic schedule analysis is performed in order to calculate

the deterministic duration, determine the critical path and non-critical activities. This is

followed by performing the stochastic analysis through the use of Monte Carlo

Simulation (MCS) for 10,000 runs. The mean project duration and its standard deviation

are then determined. The probability of finishing the project within the deterministic

duration is also calculated.

Stage 2- Schedule compression analysis

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Schedule compression analysis is performed through several cycles of crashing until the

optimum duration (associated with the minimum total cost) is reached. The number of

cycles depends on the project’s size and complexity. During each crashing cycle, all

critical activities are evaluated to select the activity that has the least impact on project

cost. The impact on project cost includes the extra direct cost due to crashing in addition

to the extra cost associated with float loss. This means that for each crashing cycle, the

float cost needs to be calculated for each activity.

For each crashing cycle (i), there is a need to determine the activity associated

with the least impact on the total project cost. For each critical activity (j) that can be

crashed, the following steps are performed:

a) Shorten the duration of activity (j) by one day and calculate the new

deterministic duration (Ddet, ij).

b) Perform stochastic analysis over the new crashed network by running Monte

Carlo Simulation for 10,000 runs.

9
 c) Find the new mean duration (Mij) and standard deviation (Stdij). Also, find

the new probability of finishing (POFi,j) the project within the new

deterministic duration. The probability of completing the project on time

should drop due to the float loss associated with shortening the project

duration.

d) Fix the probability of finishing to the value obtained in the previous cycle

(POFi-1) and calculate the associated probabilistic project duration (Dprob,ij).

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

This duration represents the desired duration without increasing project risk.

This is calculated using the probability of finishing at the previous step and

the Mean (Mij) and Standard Deviation (Stdij) found after the new simulation

run in step (b).

e) Using Equation 1, calculate the difference between the newly obtained

probabilistic duration and the new deterministic crashed duration at this step

to find the float loss impact (FLD) in terms of days:

FLDij= Dprob,ij – Ddet, ij (1)

Where;

Dprob,ij = Probabilistic duration for crashed activity j at cycle i;

Ddet, ij = deterministic duration for crashed activity j at cycle i;

FLDij = Float loss impact in terms of days for crashed activity j at cycle i.

f) Calculate the Float Loss Cost (FLC) using Equation 2. Equation 2 represents

the float loss cost as a product of the duration difference and the cost

(savings) per day.

10
 FLCij = FLDij x CSPD (2)

Where;

CSPD = Savings per day (indirect cost, incentives, etc.);

FLCij = Float loss cost for crashed activity j at cycle i.

g) Add the float cost to the extra direct cost to find the new total extra cost

h) Steps (a) through (g) are repeated for all identified critical activities. The

activity with the least total extra cost will be crashed at this point.
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

i) Float impact in days (at the following crashing cycle) is determined by

subtracting the duration associated with probability of finishing the project

on time of the previous cycle minus the new deterministic duration of the

new crashed schedule at the current cycle.

j) Steps (a) through (i) are repeated progressively until reaching the optimum

solution

If two or more critical paths are available, the following steps must be undertaken:

a) Check the available activities to be crashed (either a common activity or two

or more activities that correspond to the lowest crashing slope)

b) Perform the same steps that were performed when one critical path has

occurred while considering all possible cases occurring, then compare the

accepted cases to select the activity/activities with the lowest crashing

impact

11
 Application example project

An example project (adopted from Hinze (2012)) is used to explain the proposed

framework. The first part explains the deterministic approach to compute the optimum

project duration and total cost and the minimum project duration and its associated total

cost. Three cycles are needed to reach the optimum project duration, while two extra

cycles are needed to reach the least project duration. The optimum project duration is 23

days with an associated cost of $12,490, while the minimum project duration and its
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

associated cost are 20 days, $12,600, respectively.

Table I presents the project data in terms of durations and costs. The indirect

cost is assumed to be $280 per day. The last two columns; duration mean and duration

standard deviation, are added to Hinze (2012) example in order to use them in the

stochastic analysis.

<Insert Table I here>

Normal project compression (deterministic approach)

This section presents the solution cycles using normal deterministic project compression

without considering the effect of float loss. Based on the baseline schedule network

represented in Figure 2, the project duration is found to be 27 days, with an associated

total project cost of $12,860 that consists of a direct cost of $5,300 and indirect cost of

$7,560. The critical path is A, B, F, H, and K.

<Insert Figure 2 here>

12
 For the first crashing cycle (cycle 1), the least expensive activity to expedite is F;

therefore, the decision is to expedite F by 2 days. The new project duration is 25 days

and the new critical path is A, B, F, H, K. The new total cost is calculated as follows:

• New Direct Cost = Direct Cost + Crashing Cost = 5,300 + (2*150) = $5,600

• New Indirect Cost = Duration * indirect cost per day= 25 * 280 = $7,000

• New Total Cost = Direct Cost + Indirect Cost = 5,600 + 7,000 = $12,600
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

The updated schedule at step two is represented in Figure 3.

<Insert Figure 3 here>

In cycle 2, activity B is crashed by 1 day. In cycle 3, activity H is crashed by 1 day. In

cycle 4, activities B& C are crashed by 2 days each. In cycle 5, activity K is crashed by

1 day. Table II illustrates the crashing results performed over cycles zero to five.

<Insert Table II here>

Figure 4 illustrates the total project cost vs. duration curve. It can be noticed that the

optimum project duration and total cost are 23 days, $12,490 respectively. Afterwards,

the total cost starts to increase until reaching the cost associated with minimum project

duration of 20 days.

<Insert Figure 4 here>

Table III presents the total float available for noncritical activities at the end of each

cycle. Table III clearly shows the float loss associated with time-cost trade-off. For

example, the available float for activity D has dropped from 7 days to 1 day in cycle 5.

This reduces the schedule flexibility and reduces the chances of completing the project

on time.

13
 <Insert Table III here>

Project compression considering float consumption impact

This section presents the solution cycles using the new proposed stochastic compression

framework considering the impact of float loss within noncritical activities during the

crashing process. The normal schedule is shown in Figure 2. The project duration of the

baseline schedule is 27 days, with a direct cost of $5,300, indirect cost of $7,560, and
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

total cost $12,860. The critical path is A, B, F, H, and K. Using Monte Carlo simulation

for 10,000 runs on the normal schedule, the mean duration is calculated to be 28.1 days

with a standard deviation of 3.47 days. The probability of completing the project within

the deterministic duration of 27 days is calculated to be 37.47%.

For cycle 1, the activities available for crashing are: B, F, H, K (activity A can’t

be expedited). Therefore, four crashing scenarios need to be checked to find the best

activity to crash. The first scenario is for crashing activity B by 1 day. After crashing

activity B by 1 day, Monte Carlo Simulation is performed on the crashed schedule. The

resulting mean and standard deviation after crashing activity “B” are 27.5 days and 3.4

days respectively. The probability of completing the project on time dropped to 32.93%.

This drop is due to the loss of float after crashing activity “B”. To calculate the effect of

the float loss and determine the new project cost, the following steps are followed:

New deterministic duration Ddet, 1B = 26 days

Duration at 37.75% given the new mean and standard deviation is

(Dprob,1B) = 26.42 days

According to Equation 1, the difference between the deterministic and the new

duration: FLD1B = Dprob,1B – Ddet, 1B =26.42 – 26 = 0.42 days

14
 Float cost FLC1B is calculated according to Equation 2:

FLC1B = FLD1B x CSPD = 0.42 (days) x 280 $ / day = $116.70

Extra direct cost (slope) = $ 200

Total extra cost = 200 + 116.70 = $316.70

Direct Cost Now = total extra + direct cost = $5,616.70

Indirect Cost = $7,280

Total Cost = $12,896.70

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

The same steps are followed for the other three scenarios namely crashing activities

“F”, “H” and “K” by one day each separately. The results are summarized in Table IV.

<Insert Table IV here>

According to the total extra cost and total cost, activity F exhibited the least total extra

cost and total cost. Therefore, the decision is to expedite activity F by 1 day.

For cycle 2, the activities available to be expedited are: B, F, H, and K. Based on that,

four available scenarios need to be checked to find the best activity to crash. Table V

summarizes the simulation results for cycle 2.

<Insert Table V here>

According to the total extra cost and total project cost, activity F exhibited the least total

extra cost and total project cost. Based on that, the decision in this cycle is to expedite F

by 1 day. For cycle 3, three available scenarios have to be checked to find the best

activity to crash at cycle three. The three scenarios include either crashing activity B or

H or K. Table VI summarizes the simulation results for cycle 3.

15
 <Insert Table VI here>

The best activity to crash in cycle 3 is activity “B”. However, the total Cost started to

increase at this cycle; therefore, the optimum project duration and total cost are 25 days

and $12,709 respectively.

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Analysis and discussion of the results

As per the example presented and solved earlier, the optimum duration considering float

consumption impact is 25 days, while the associated optimum total cost is found to be

$12,709. Table VII compares the remaining total float for the noncritical activities

between the deterministic compression method and the new proposed compression

framework.

<Insert Table VII here>

It can be noticed that the new proposed compression model is better in terms of

remaining float as it finds an optimum solution that can save some total float for future

use. In terms of the probability of finishing the project on time, the probability of

finishing the project within 27 days is found to be 0.375. The probability of finishing

the project within 25 days when float loss impact is considered is 0.331, while the

probability of finishing the project within 23 days when float loss impact is not

considered is 0.237. It can be seen that when float loss impact is considered, the

probability of finishing the project is considerably higher than that when float loss

impact is not considered. Figure 5 compares the results between the optimum solution

16
 found using deterministic approach (without float loss effect) and the optimum solution

found considering the float loss effect in terms of total cost curves.

<Insert Figure 5 here>

From Figure 5, it is clear that the optimum project total cost considering float

consumption impact is higher than the optimum normal cost. The optimum project

duration, as well, is higher than that when float loss cost isn’t considered. The increase

in the project’s total cost of the curve considering float consumption impact is related to
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

the increase in the direct cost that accounts for the float loss cost in noncritical

activities. The framework presents a curve with a higher cost. The difference between

the optimum total cost when float is considered and the deterministic normal optimum

total cost is equal to $ 219, in this example. This higher cost, if paid, accounts for and

quantifies the float cost impact and can save dollars associated with risks resulting from

the loss of project flexibility. Decision makers or project managers, depending on the

nature of their projects, are free to choose between the two curves; whether to stick to

the normal compression method and bear the risk associated with losing total float, or

use the new curve and be on the safe side while maintaining a compressed schedule.

To best check the applicability of the developed framework, the framework was

tested against five examples selected from literature. The probabilities of finishing the

project on time using the developed framework in all five cases were better than those

using the classical deterministic optimization technique. Table VIII summarizes the

results of the five examples.

<Insert Table VIII here>

17
 Summary and conclusions

The objective of time-cost trade-off is to determine the optimum project duration

corresponding to the minimum total cost. Time-cost trade-off techniques result in

reducing the available float for noncritical activities and thus increasing the schedule

risks. Existing deterministic optimization technique doesn’t consider the impact of the

float loss within the noncritical activities when the project duration is being crashed.
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

The proposed framework incorporates float loss impact into the time-cost trade-

off problem. The float loss impact is quantified using Monte Carlo Simulation by

determining the probability of finishing the project at each crashing cycle for each

activity and comparing the total extra direct cost (activity cost slope and float loss cost)

to select the activity that best exhibits the least total cost. As the project duration is

reduced, the available float of non-critical activities is reduced, and the probability of

finishing the project on time is also reduced. The framework measures the difference

between the probabilistic duration and the deterministic duration and translates this loss

of time into a cost that is added to the project direct cost.

The stochastic framework uses Monte Carlo Simulation to calculate the effect of

float loss on risk. This is later translated into an added cost to the trade-off problem.

Five examples, from literature, are solved using the proposed framework in order to test

the applicability of the developed framework. The results confirmed the research

hypothesis that the new optimum solution will be at a higher duration and cost but at a

lower risk compared to traditional methods. The probabilities of finishing the project on

time using the developed framework in all five cases were better than those using the

18
 classical deterministic optimization technique. The results indicate that the proposed

framework reduces the risks associated with float loss. As seen in the example, the

probability of finishing the project on time was considerably higher when float loss

impact was considered. Although the optimum duration and cost are higher than those

obtained from traditional methods, the schedule risk is lower. The proposed framework

provides decision makers with a new tool to solve the time-cost trade-off problem with

the least possible risks associated with float loss. The proposed framework allows
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

project managers to determine the optimum project duration at the least cost and risk.

The traditional methods allow project managers to determine the optimum project

duration at the least cost but at a higher risk levels. So, the proposed framework allows

project managers to experience a three-way trade off between time, cost and risk. They

may choose the project duration associated with the minimum cost (ignoring float loss

cost) or they may choose a higher duration with a higher cost (higher than the optimum

duration and minimum cost, respectively) but with less risk.

This paper introduces a new concept to the time cost trade-off problem, which is

the incorporation of float loss impact. Previous researches on time cost trade off ignored

the effect of float loss that results from shortening the project duration. This research

attempts to fill that gap in literature. Future research may target developing other

optimization methods or meta heuristic techniques to speed up the calculation process.

19
 Appendix A – Notation

i: Crashing cycle (1….n); n: number of crashing cycles;

j: Activity (1….k); k: number of activities in the project;

CSPD: Savings per day (indirect cost, incentives, etc.);

Ddet, ij: Deterministic duration for crashed activity j at cycle i;

Dprob,ij: Probabilistic duration for crashed activity j at cycle i;

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

FLCij: Float loss cost for crashed activity j at cycle i;

FLDij: Float loss impact in terms of days for crashed activity j at cycle i;

Mij: Mean project duration for cycle i after crashing activity j;

POFi-1: Probability of finishing on time for the previous cycle;

POFij: Probability of finishing on time for cycle i after crashing activity j;

STDij: Project duration standard deviation for cycle i after crashing activity j.

20
 References

Al-Gahtani, K. (2009), "Float allocation using the total risk approach", Journal of
Construction Engineering & Management, Vol. 135, No. 2, pp. 88-95.

Al-Gahtani, K. and Mohan, S. (2005), “Total float management for delay analysis”,
AACE International Transactions, CDR.16.1-13.

Afshar, A., Ziaraty, A., Kaveh, A. and Sharifi, F. (2009), “Non-dominated archiving
multi-colony ant algorithm in time–cost trade-off optimization”, Journal of
Construction Engineering and Management, Vol. 135, No. 7, pp. 668-674.

Al Haj, R. and El-Sayegh, S. (2015), “Time-cost optimization model considering float-

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

consumption impact”, Journal of Construction Engineering and Management, Vol. 141,

No. 5, 10.1061/(ASCE)CO.1943-7862.0000966, 04015001.

Ammar, M. (2003), “Float analysis of non-serial repetitive activities”, Construction

Management and Economics, Vol. 21, No. 5, pp. 535-542.

Ammar, M. (2011), “Optimization of project time-cost trade-off problem with

discounted cash flows”, Journal of Construction Engineering and Management, Vol.
137, No. 1, pp. 65-71.

Castro-Lacouture, D., Süer, G., Gonzalez-Joaqui, J., and Yates, J. (2009), “Construction
project scheduling with time, cost and material restrictions using fuzzy mathematical
models and critical path method”, Journal of Construction Engineering & Management,
Vol. 135, No.10, pp. 1096-1104.

Charoenngam, C. and Popescu, C. (1995), Project Planning, Scheduling and Control in

Construction: An Encyclopedia of Terms and Applications, 1st Ed. New York, US: John
Wiley & Sons Press.

Chassiakos, A. and Sakellaropoulos, S. (2005), “Time-cost optimization of construction

projects with generalized activity constraints”, Journal of Construction Engineering and
Management, Vol. 131, No. 10, pp. 1115-1124.

Choudhry, R., Aslam, M., Hinze, J., and Arain, F. (2014). "Cost and Schedule Risk
Analysis of Bridge Construction in Pakistan: Establishing Risk Guidelines", J. Constr.
Eng. Manage., Vol. 140, No. 7, 10.1061/(ASCE)CO.1943-7862.0000857, 04014020.

De la Garza, J., Prateapusanond, A. and Ambani, N. (2007), “Preallocation of total float

in the application of a critical path method based construction contract”, Journal of
Construction Engineering & Management, Vol. 133, No. 11, pp. 836-845.

21
 De la Garza, J., Vorster, M. and Parvin, M. (1991), “Total float traded as commodity”,
Journal of Construction Engineering and Management, Vol. 117, No. 4, pp. 716-727.

Deckro, R., Herbert, J., Verdini, W., Grimsrud, P. and Venkateshwar, S. (1995),
“Nonlinear time-cost trade-off models in project management”, Computers ind. Eng,
Vol. 28, No. 2, pp. 219-229.

El-Sayegh, S. and Mansour, M. (2015), “Risk Assessment and Allocation in Highway

Construction Projects in the UAE”, Journal of Management in Engineering, Vol. 31,
No. 6, 10.1061/(ASCE)ME.1943-5479.0000365, 04015004.

Elbeltagi, E. (2002), “Project time-cost trade-off”, Construction Management, pp. 164-

181. http://osp.mans.edu.eg/elbeltagi/CM%20CH8%20Time-Cost.pdf
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Eshtehardian, E., Afshar, A. and Abbasnia, R. (2008), “Time-cost optimization: using

GA and fuzzy sets theory for uncertainties in cost”, Construction Management and
Economics, Vol. 26, No. 7, pp. 679-691.

Feng, C., Liu, L. and Burns, S. (2000), “Stochastic construction time-cost trade-off
analysis”, Journal of Computing in Civil Engineering, Vol. 14, No. 2, pp. 117-126.

Geem, Z. W. (2010), “Multiobjective optimization of time-cost trade-off using harmony

search”, Journal of Construction Engineering and Management, Vol. 136, No. 6, pp.
711-716.

Gong, D. (1997), “Optimization of float use in risk analysis-based network scheduling”,

International Journal of project Management, Vol. 15, No. 3, pp. 187-1.

Gong, D. and Rowings, J. (1995), "Calculation of safe float use in risk-analysis-oriented

network scheduling", International Journal of Project Management, Vol. 13, No. 3, pp.
187-194.

Gould, F. E. (2005), Managing the Construction Process: Estimating, Scheduling and

Project Control, 3rd Ed., New Jersey, US: Prentice Hall.

Hazir, O., Erel, E. and Gnalay, Y. (2011), “Robust optimization models for the discrete
time/cost trade-off problem”, International Journal of Production Economics, Vol. 130,
No. 1, pp. 87-95.

Hegazy, T. and Menesi, W. (2012), “Heuristic method for satisfying both deadlines and
resource constraints”, Journal of Construction Engineering and Management, Vol. 138,
No. 6, pp. 688-696.Hinze, J. W. (2012), Construction Planning & Scheduling, 4th ed.,
New Jersey, US: Pearson Prentice Hall.

22
 Householder, J. and Rutland, H. (1990), "Who owns float?" Journal of Construction
Engineering & Management, Vol. 116, No. 1, pp. 130-133.

Isidore, L. and Back, W. E. (2001), “Probabilistic optimal-cost scheduling”, Journal of

Construction Engineering & Management, Vol. 127, No. 6, pp. 431-437.

Islam, M. N., Rana, M. B., Rafique S., and Aziza T. (2004), “Crashing project time with
least cost: a linear programming approach”, Journal of business research, Vol. 6, pp.
13-28.

klansek, U. and Psunder, M. (2008), “Cost optimal project scheduling”, Journal of

Management, Information Systems & Human Resources, Vol. 41, No. 4, pp. 153-158.
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Larson, E. and Gray, C. (2014), Project Management: the Managerial Process, 6th ed.,
New York, US: McGraw-Hill Press.

Leu, S. & Yang, C. (1999), “GA-based multi-criteria optimal model for construction
scheduling”, Journal of Construction Engineering & Management, Vol. 125, No. 6, pp.
420-427.

Li, H. & Love, P. (1997), “Using improved genetic algorithms to facilitate time-cost
optimization”, Journal of Construction Engineering & Management, Vol. 123, No. 3,
pp. 233-237, 1997.

Liu, J. and Rahbar, F. (2004), “Project time-cost trade-off optimization by maximal

flow theory”, Journal of Construction Engineering & Management, Vol. 130, No. 4, pp.
607-609.

Liu, L., Burns, S., and Feng, C. (1995), “Construction time-cost trade-off analysis using
LP/IP hybrid method”, Journal of Construction Engineering & Management, Vol. 121,
No. 4, pp. 446-454.

Moussourakis, J. and Haksever, C. (2007), “Models for accurate computation of earliest

and latest start times and optimal compression in project networks”, Journal of
Construction Engineering & Management, Vol. 133, No. 8, pp. 600-608.

Ng., S. and Zhang, Y. (2008), “Optimizing construction time and cost using ant colony
optimization approach”, Journal of construction Engineering & Management, Vol. 134,
No. 9, pp. 721-728.

Oxley, R. and Poskitt, J. (1996), Management Techniques Applied to the Construction

Industry, 5th ed., Oxford: Blackwell Science Ltd.

Perera, S. (1980), “Linear programming solution to network compression”, Journal of

Construction Division, Am. Soc. Civ. Eng., Vol. 106, No. 3, pp. 315-326.

23
 PMBOK (2013), A Guide to the Project Management Body of Knowledge, Project
Management Institute Standards Committee, PMI.

Ponce de Leon, G. (1986), "Float ownership: Specs treatment", Cost Engineering, Vol.
28, No. 10, pp. 12-15.

Que, B. (2001), “Incorporating practicability into genetic algorithm-based time-cost

optimization”, Journal of Construction Engineering & Management, Vol. 128, No. 2,
pp. 139-143.

Sakka, Z. and El-Sayegh, S. (2007), “Float consumption impact on cost and schedule in
the construction industry”, Journal of Construction Engineering & Management, Vol.
133, No. 2, pp. 124-130.
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Sonmez, R. and Bettemir, O. H. (2012), “A hybrid genetic algorithm for the discrete
time-cost trade-off problem”, Expert Systems with Applications, Vol. 39, No. 13, pp.
11428-11434.

Wickwire, J., Warner, T. and Berry, M. (1999), "Construction Scheduling", in

Construction Law handbook. Gaithersburg, N. Y.: R. F. Cushman and J. J. Myers, eds.,
Aspen Law & Business.

Yang, I-T. (2005), “Chance-constrained time-cost trade-off analysis considering

funding variability”, Journal of Construction Engineering & Management, Vol. 131,
No. 9, pp. 1002-10.

Yang, I-T. (2007), “Using elitist particle swarm optimization to facilitate bi-criterion
time-cost trade-off analysis”, Journal of Construction Engineering and Management,
Vol. 133, No. 7, pp. 498-505.

Zahraie, B. and Tavakolan, M. (2009), “Stochastic time-cost-resource utilization

optimization using non-dominated sorting genetic algorithm and discrete fuzzy sets”,
Journal of Construction Engineering and Management, Vol. 135, No. 11, pp. 1162-
1171.

Zeinalzadeh, A. (2011), “An application of mathematical model to time-cost trade-off

problem (case study)”, Australian Journal of basic & Applied Sciences, Vol. 5, No. 7,
pp. 208-214.

Zhang, H. and Li, H. (2010), “Multi-objective particle swarm optimization for

construction time-cost trade-off problem”, Construction Management and Economics,
Vol. 28, No. 1, pp. 75-88.

24
 Zhang, Y. (2016), “Selecting risk response strategies considering project risk
interdependence”, International Journal of Project Management, Vol. 34, No. 5, pp
819-830.

Zheng, D. and Ng, T. (2005), “Stochastic time-cost optimization model incorporating

fuzzy sets theory and non-replaceable front”, Journal of Construction Engineering and
Management, Vol. 131, No. 2, pp. 176-186.

Ziegler, H. (1985), "Minimal and maximal floats in project networks", Engineering

Costs and Production Economics, Vol. 9, pp. 91-97.
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

About the authors

Sameh El-Sayegh is an Associate Professor of Civil Engineering at the

American University of Sharjah. He has an M.S. degree in Construction Management

and a Ph.D. degree in construction Engineering and Project Management from Texas

A&M University, College Station, Texas – USA. He worked in project controls with

Consolidated International Contracting Company (CCIC) in Greece, Yemen and USA.

He is a certified Project Management Professional (PMP) from Project Management

Institute (PMI).

Rana Al-Haj is a Civil Engineer and a LEED Green Associate Professional. She

has a M.S. degree and a BS degree in Civil Engineering from the American University

of Sharjah with a minor in Engineering Management and Construction Management.

She is currently working in the Construction Management and Project Controls

Department at Dar AL- Handasah Consultants (Shair and Partners) over Dubai

International Airport and Al Maktoom International Airport projects.

25
 - Precedence Diagram
- PDF* of Activities Durations

Stochastic Analysis using MCS*

- Normal Project Duration Mij* & Stdij*

- Normal Project POF*

Identify Critical Activities

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Crash Critical Activity / Activities by 1 day

Stochastic Analysis using MCS

- Find New Project Mij & Stdij

- Find New Probability of Finishing POF
-Find Ddet. ij* and Dprob. ij* Redo for
other
Critical
Activities

Calculate FLDij * = Dprob. ij – Ddet. Ij

FLCij*= FLDij* x CSPD*

Find TDC* = Sj* + FLCij

And TC* = TDC+ IDC*

Crash Activity / Activities with Least

TDC/TC

Optimum
Stop Crashing YES Solution is NO Continue Crashing
reached?

PDF: Probability Distribution Function MCS: Monte Carlo Simulation Mij: Schedule Mean @ cycle i when activity j is crashed
Stdij: Schedule Standard Deviation @ cycle i when activity j is crashed POF: Probability of Finishing on Time
Ddet. ij: Deterministic Duration for Crashed Activity j at cycle i Dprob. ij: Probabilistic Duration for Crashed Activity j at Cycle i FLDij:
Duration Difference Between Dprob. ij Associated with the POF at the Crashing Cycle in Question & Ddet. ij Sj: Activity j Crashing Slope FLCij:
Float Loss Cost CSPD: Savings Per Day TC: Total Cost TDC: Total Direct Cost IDC: Indirect Cost
 Legend ES Dur EF
Act TF
LS act DC LF

Cycle 0
Direct Cost 5300
Indirect Cost 7560
Total Cost 12860
Duration 27 days
C.P. A,B,F,H,K
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

1 7 8 8 3 11 15 7 22
B 0 E 4 H 0
1 1000 8 12 100 15 15 350 22

0 1 1 1 6 7 8 7 15 15 5 20 22 5 27
A 0 C 1 F 0 I 2 K 0
0 800 1 2 300 8 8 500 15 17 700 22 22 450 27

1 3 4 4 8 12 15 3 18
D 7 G 7 J 4
8 400 11 11 200 19 19 500 22

Figure 2 Cycle zero – normal schedule

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Figure 3
Cycle one – crashed schedule 1
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Figure 4
Total cost vs. duration
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Figure 5
Total cost vs. duration comparison
 Table I Example project – normal and crash cost & duration

Activity Normal Normal Crashed Crashing Potential Cost Duration Duration

Duration Cost Duration Cost Days per Mean Standard
Saved Day Deviation
A 1 800 1 800 0 - 1 1.2
B 7 1,000 4 1,600 3 200 7 2
C 6 300 4 500 2 100 6 1.5
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

D 3 400 2 800 1 400 3 1.35

E 3 100 1 200 2 50 3 1.88
F 7 500 5 800 2 150 7 2.12
G 8 200 4 1,400 4 300 8 3
H 7 350 6 600 1 250 7 1.25
I 5 700 3 850 2 75 5 2.5
J 3 500 2 1,000 1 500 3 1.5
K 5 450 4 800 1 350 5 1.6
Total= 5300
 Table II Project crashing results

Cycle Project Duration Direct Cost Indirect Cost Total Cost

0 27 5,300 7,560 12,860
1 25 5,600 7,000 12,600
2 24 5,800 6,720 12,520
3 23 6,050 6,440 12,490
4 21 6,650 5,880 12,530
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

5 20 7,000 5,600 12,600

 Table III Activities’ total float (days) at each cycle

Activity Cycle 0 Cycle 1 Cycle 2 Cycle 3 Cycle 4 Cycle 5

C 1 1 0 0 0 0
D 7 5 4 3 1 1
E 4 2 2 2 2 2
G 7 5 4 3 1 1
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

I 2 2 2 1 1 1
J 4 4 4 3 1 1
 Table IV Simulation results for cycle 1

B F H K

Mean (days) 27.5028 27.16 27.4942 27.1397

Standard Deviation (days) 3.401 3.3939 3.6883 3.4861

Probability of finishing within

32.9291627% 36.6254588 % 34.2695061% 37.1861573%
26 days (%)

Duration at 37.74742% 26.41679274 26.07625991 26.31645211 26.02651861

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Float Loss Impact (days) 0.41679274 0.07625991 0.31645211 0.02651861

Float Loss Cost ($) $116.7 $21.35 $88.61 $7.43

Total Extra Cost ($) $316.7 $171.35 $338.61 $357.43

New Direct Cost ($) $5,616.7 $5,471.35 $5,638.61 $5,657.43

Indirect Cost ($) $7,280 $7,280 $7,280 $7,280

Total Cost ($) $12,896.7 $12,751.35 $12,918.61 $12,937.43

 Table V Simulation results for cycle 2

B F H K

Mean (days) 26.591 26.445 26.5491 26.2251

Standard Deviation (days) 3.3549 3.3059 3.3519 3.4652

Probability of finishing within

31.7667338% 33.1020389% 32.1984801% 36.1840397%
25 days (%)

Duration at 36.6254588% 25.4443298 25.31507749 25.40345517 25.0407304

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Float Loss Impact (days) 0.4443298 0.31507749 0.40345517 0.0407304

Float Loss Cost ($) $124.41 $88.22 $112.97 $11.40

Total Extra Cost ($) $324.41 $238.22 $362.97 $361.40

New Direct Cost ($) $5,795.77 $5,709.57 $5,834.32 $5,832.76

Indirect Cost ($) $7,000 $7,000 $7,000 $7,000

Total Cost ($) $12,795.77 $12,709.57 $12,834.32 $12,832.76

 Table VI Simulation results for cycle 3

B H K

Mean (days) 25.106 25.8742 25.6759

Standard Deviation (days) 3.2834 3.3498 3.5006

Probability of finishing
28.0318093% 28.79116% 31.605949%
within 24 days (%)

Duration at 33.1020389% 24.47543469 24.41001143 24.14579715

Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Float Loss Impact (days) 0.47543469 0.41001143 0.14579715

Float Loss Cost ($) $133.12 $114.80 $40.82

Total Extra Cost ($) $333.12 $364.80 $390.82

New Direct Cost ($) $6,042.70 $6,074.38 $6,100.40

Indirect Cost ($) $6,720 $6,720 $6,720

Total Cost ($) $12,762.70 $12,794.38 $12,820.40

 Table VII Comparison of the remaining total float

Noncritical Activity Activity Total Float in

Activity Total Float in Days @ 25 Days Duration
Days @ 23 Days Duration (New Proposed
(Deterministic) Framework)

C 0 1
D 3 5
E 2 2
G 3 5
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

I 1 2
J 3 4
 Table VIII Summary of the results of the tested five examples

Manual- Probabilistic Framework

Normal Case Deterministic Solution
Solution
Example
Project Total Project Total Project Total Project
POF POF POF
Duration Project Cost Duration Project Cost Duration Cost

Example One
(Isidore & 20 days $8,215 0.355 12 days $7,940 0.109 17 days $8,146.5 0.253
Downloaded by Florida Atlantic University At 04:59 11 March 2017 (PT)

Back 2001)
Example
Two
24 days $30,520 0.426 17 days $28,870 0.209 18 days $29,742.5 0.250
(Oxley &
Poskitt 1996)
Example
Three
20 days $2,044,000 0.342 16 days $1,990,000 0.122 18 days $2,020,269.525 0.204
(Zeinalzadeh
2011)
Example
Four
59 days $43,875 0.393 51 days $43,545 0.226 54 days $ 43,792.73 0.356
(Elbeltagi
2002)

Example Five
37 days $119,745 0.342 30 days $116,465 0.165 35 days $ 116,604.992 0.338
(Gould 2005)