S13-5

AN ECONOMIC EVALUATION SYSTEM FOR BUILDING CONSTRUCTION

PROJECTS IN THE CONCEPUTAL PHASE

Hyung-Jin Kim* and Yong-Chil Seo

Department of Research & Development, Kunwon Engineering Co. Ltd., Seoul, Korea
* Corresponding author ([email protected])

ABSTRACT: The conceptual phase of capital projects is of strategic importance - an economic evaluation at this early
stage is critical to the owners’ decision-making with limited information. The current study developed a comprehensive but
practical system that can be used in such an early phase for economic evaluation of capital building construction projects.
The system integrates four analytical steps: project planning, construction costs estimation, projection of incomes and
expenditures, and economic evaluation. Due to the integration, the system effectively manages the information flow from
the planning to the evaluation, in which any changes from the initial plan are fully accounted for multiple alternates. In the
system, users predict construction cost using a combined estimating method with historical project data and assembly costs.
The system performs the tabulation of all incomes and expenditures based on the automated calculation as well as users’
inputs and then the economic evaluation based on project cash flow, which is generated using an automated scheduling
method. System performance has been tested in terms of its accuracy and efficiency through case study.

Keywords: Economic Evaluation, Capital Projects, Conceptual Phase, Automation, Historical Database

1. INTRODUCTION it between the analytical steps, especially more when

Economic evaluation of facility investment is one of the multiple alternates need to be analyzed. Also, controlling
most critical tasks in the conceptual phase since it involves the information and the changes becomes complex if a
a commitment of resources spending when a decision is project includes multiple buildings of mixed-use. The
made. The evaluation involves forecasting cash flow current study develops a comprehensive but practical
considering all capital requirements and revenues [1]. An economic evaluation system to combine the analytical
economic evaluation should include the following steps so that the information flow can be efficiently
analytical steps: 1) project planning; 2) construction cost managed in an integrated way based the computerized
estimating; 3) projection of all expenditures and revenues; automation, but using limited project information available
and 4) cash flow analysis and economic evaluation for in an early phase of project planning.
alternates [2]. However, most existing software programs
in construction have been developed separately for cost 2. EFFICIENCY OF EVALUATION SYSTEM
estimation or for project evaluation while they have The essential element of economic evaluation is project
different levels of detail in the analysis. Thus, the cash flow analysis, which requires quantifying all
efficiency of the evaluation task has been limited – it is monetary items and specifying the time of their occurrence.
difficult to manage project information and any revision of Fig. 1 conceptually depicts the information for an

467
 S13-5

economic evaluation of a capital building construction estimator selects a similar one from past projects and
project. All expenditures and revenues should be accounted adjusts it based on similarities and differences perceived by
for the analysis in forms of quantity and time within the the estimator, which depends on his individual experience
prediction period. For multiple alternates (the initial plan or perception. Also, simple square foot or assembly costs
and its revisions), quantifying and scheduling of all the methods are preferred by many cost estimators in the
items for each alternate is a challenging task. A practical conceptual phase [5]. However, under limited project
system, for the use by practitioners, should be able to deal information in the conceptual phase, lack of historical data
with all the quantity and time information in an integrated frequently could lead to the estimator’s bias, which could
way based on automated estimating and scheduling. result in under budget or over budget [6].
However, the system should consider the trade-off between
automation and flexibility – the efficient use of un-detailed 3.2. Combined Cost Estimating
project information supported by the automation, as well as Estimators usually desire to have their estimates compared
the flexibility, i.e., allowing manual adjustments by users. with the actual costs of past similar projects, which makes
them feel comfortable. Also, many professionals in

Information for
construction desire to have a list of quantities and costs in a
an Economic Evaluation
form of cost breakdown structure (CBS), in which the real
Quantity

values are present for different work packages even though

Time
the CBS has a low level of detail in an early phase. For
Quantity & Time instance, a building is decomposed into work elements

All Capital Expenditures All Expected Revenues whereas unit cost of each element can be assessed and the
- Land Acquisition - Sales total cost is the sum of the products of the quantities
- Construction - Rents
- O & M Costs
- Other Expenditures
- Operating
- Other Revenues
multiplied by their corresponding unit costs.
Considering the points addressed above, the current study
Fig. 1 Information for economic evaluation proposes a combined estimating method using historical
data and assembly costs for the conceptual phase. There are
3. CONSTRUCTION COST ESTIMATING advantages by using both methods simultaneously to
3.1. Estimating in the Conceptual Phase determine an estimate in such an early phase. A historical
In the conceptual phase, cost estimate with a reasonable data-based estimate can be made using similar projects
accuracy is required within a limited time period while data selected based on their similarity to the current project
information is limited [3]. Various estimating methods for while the similarity is determined by major project
the conceptual phase have been proposed: simple square information (design parameters). At the same time, an
foot method, regression analysis, probabilistic estimating, assembly cost-based estimate can be prepared by
neural network, case-based reasoning, fuzzy models, etc. calculating the required quantity and cost of an assembly (a
Meanwhile there are criticisms on some empirical set of material/labor/equipment) for major construction
estimating models, e.g., black box with no explanation of methods selected by estimators for different work packages.
output [4]. In the conceptual phase, cost estimators Even if using such un-detailed information, the comparison
commonly use analogy-based estimating methods, i.e., an

468
 S13-5

of the two estimates can reduce the chances of possible different equations. The accuracy of the output can be
under or over estimate and improve estimators’ confidence. improved as more data are accumulated in the system.

3.3. Data for the Combined Cost Estimating 4. DEVELOPMENT OF THE SYSTEM
The data for the estimating method described above 4.1. Framework of the System
include historical data for past projects and different types The system consists of four analytical steps (project
of assembly costs. The historical data, in which the projects planning, construction costs estimation, projection of
(all single-use buildings) are classified by the type of incomes and expenditures, and economic evaluation) and
building, provide unit cost as well as duration information. database. The database in the current system includes
Different types of building should be considered in cost historical data (cost and duration) for 301 past projects and
estimating since the type of building has an important 131 different types of assembly (equations and composite
effect on the project cost [7]. Each project needs to be unit costs). Fig. 2 shows the overall framework of the
indexed by major parameters, i.e., site, size of building system, in which project information flows from Project
(square foot and story), building-to-land ratio, floor space Planning to Economic Evaluation. The system is dynamic
index, etc. These building characteristics are used as the in terms of that changes in a previous step are
multiple search conditions to find similar projects to automatically reflected in the next steps. Commonly,
estimate the current project. When a building includes project planning and cost estimating are performed using
multiple uses, the system calculates a weighted average estimating software programs while the others separately
unit cost per area (M2) to estimate the total cost whereas using economic evaluation software programs. A detailed
weights are determined based on the areas of different uses. description of each analytical step and linkage between the
An assembly represents a specific construction method, steps is provided below.
which involves a set of materials, labors, and equipments.
Each assembly has its own pre-determined equations 4.2. Project Planning
(developed by experienced estimators) to estimate the In the system, users develop a project plan for one or
required quantity for a set of materials, labors, and multiple buildings by specifying major project information,
equipments and its composite unit cost. For instance, if a which include project name, location, site area, building
user selects concrete for the superstructure of a building, area, floor area, gross floor area, building-to-land ratio,
the program estimates the quantities of materials (concrete, floor space index, number of stories, average height, areas
form, steel, etc.), labors (formwork, reinforcing, and of different uses, perimeter of building, area of exterior
concrete placement, finishing), and equipments (concrete finish, parking area, etc. Different uses in a building are
pumps) based on the design parameters, i.e., the perimeter specified by quantifying the areas of different uses. All of
of the building, stories, floor area, height, etc. the above information should be available in a schematic
The historical project data as well as assembly costs can be plan. The above building characteristics are used as
accumulated or updated by the system administrator, i.e., multiple search conditions to find similar projects from the
adding/deleting/modifying the assembly items and their historical database.
equations – different organizations (estimators) may have

469
 S13-5

- Ownership required quantities of a set of material/labor/equipment and

Planning
- Number of Buildings
Project
- Description of Building
∙ Type and uses (single or mixed) estimating the assembly cost using composite unit costs.
∙ Design parameters (area, story, building-to-
land ratio, floor space index, etc.) Even if using un-detailed information (major design
parameters and construction methods), the comparison of
Costs Estimate

- Cost Estimates
Construction

Similarity ?
∙ Estimates using historical cost data
∙ Estimates using assembly costs the two estimates improves the accuracy as well as the
∙ Comparison and adjustment by users
- Estimate of the Total Costs
- Estimate of Duration estimators’ confidence in the final estimate whereas users
may select one between them or modify depending on their
- Setting Milestones
Expenditures
Incomes &

- All Expenditures
∙ Land, design, engineering, procurement
subjective judgment.
∙ construction, operation/maintenance, taxes
- All Incomes: revenues, sales, rents
- Automatic Tabulation and Scheduling
4.4. Construction Duration
- Cash Flows (Tables & Charts)
The system predicts construction duration for each building
Evaluation
Economic

∙ Expenditures/incomes/net/cumulative
- Statistical Analysis
∙ Monte Carlo simulation / sensitivity analysis
- Comparision of Alternates
by averaging the durations of the similar projects, which
- Summary of Results
were selected in Construction Cost Estimate. When a
Database
building includes different uses, its duration per unit area
- Assembly Costs
∙ Construction method (M2) is estimated as a weighted average while the weights
∙ Material/labor/eqip.

Reports are determined based on the areas of different uses.

- Historical Cost Data
∙ Building types and uses However, users may specify the target duration if necessary.
∙ Design parameters
This duration information is delivered to the automated
Fig. 2 Framework of the system scheduling using a generic adjustable S-curve, which is
discussed below in detail.
4.3. Construction Costs Estimate
The system searches similar projects using the multiple 4.5. Expenditures and Incomes
search conditions defined in Project Planning. Similarity Total project expenditures include land acquisition,
of a past project to the current project is simply defined as architectural and engineering design, construction, field
true (matched = 1) or false (not-matched = 0) for a search supervision, financing costs, insurance and taxes,
condition and the sum of similarity scores is given to the inspection and testing, operation and maintenance costs,
project. To estimate the construction cost of the current repair costs, utilities, and other expenses. It should be
project, users select projects among the scored projects noted that construction cost is one of the cost elements of
which are prioritized by the system based on the similarity the total capital cost and other cost elements are not
score. The users’ selection of the similar projects represents insignificant [2]. The construction cost estimate is
the implementation of their experience and subjective frequently used as the reference for estimating of other
judgment. To account for the variability of cost due to various expenditures. The system provides the tabulation of
inflation, adjustments are made whenever historical data a total forty nine expenditure items (including most of
are retrieved for the estimation. legally required expenditures in Korea), which are
In addition, users are able to estimate costs for different automatically calculated using pre-determined equations
work packages using the assembly costs by selecting major (developed by experienced professionals), and different
construction methods. This method involves calculating the types of revenues defined by users. Users may add or

470
 S13-5

delete individual items and also make changes in the 4.7. Economic Evaluation
estimates if necessary. Any changes by users are reflected The current system develops four types of project cash
in the cash flow analysis. flows: expenditure, income, net, and cumulative net. Also,
four evaluation metrics are calculated: rate of return
4.6. Automated Scheduling (ROR); net present value (NPV); internal rate of return
In the system, users define major milestones within the (IRR); and payback period (PP). User defined discount rate
prediction period, such as land acquisition, beginning of is used for the evaluation of a long-term project.
construction, completion of construction, initiation of sales
or rents, etc. Fig. 3 shows the two automated scheduling 4.8. Dealing with Uncertainty
methods: 1) scheduling of construction expenditures using It is a difficult task to forecast the costs and benefits before
a generic adjustable S-curve and 2) scheduling of their realization. Information available in the conceptual
individual monetary items using one or two milestones. phase is limited as well as uncertain. All estimates of costs
The construction expenditures are planned using a generic and benefits involve some degree of uncertainty. Users
S-curve, by converting it to a density function. The curve is may execute Monte Carlo simulation for the economic
the average of percentage complete of all historical evaluation by specifying range for the estimates with
projects in the database. This curve can be prolonged or probability distributions. The sensitivity of each monetary
shortened based on the estimated construction duration for item is estimated through the simulation.
the current project. On the other hand, each of the other
monetary items (expenses and incomes) has its own default 4.9. Comparison of Multiple Alternates
sequence in the system being defined relatively to one or In the conceptual phase, multiple alternates are commonly
two major milestones. For instance, an expense is expected analyzed and compared to select the best economically
to occur at a specific time before or after one of the major feasible one. In the system, users are able to generate
milestones or periodically with an interval between two multiple copies of the initial plan, which are called
milestones. Based on the default sequences and the revisions. In each revision, users may modify project plan
projected quantities in Incomes and Expenditures, the (i.e., design parameters), as well as certain cost or revenue
system generates the project cash flow. However, users items in terms of monetary quantity as well as sequences in
may adjust the schedule for individual items. schedule. Changes made in each revision are automatically
reflected to the subsequent analyses by taking advantages
[Scheduling of Construction Costs] [Scheduling using Milestones] of the integrated information flow. As a result, the system
-2 provides an efficient comparison of different revisions
+ 1.5
Complete

-1 against the initial plan by the evaluation metrics, as well as

Interval by the probability distribution of rate of return.
Duration
Construction

5. CASE STUDY
Milestones
Prediction Period
The system has been tested for its performance. During the
Fig. 3 Automated Scheduling test, the system was applied to an urban development
project in Seoul. The project includes six buildings of

471
 S13-5

mixed-use. At the time of the test, the project already had ACKNOWLEDGEMENT
the result of feasibility study and a detailed estimate after This research was supported by a grant (07 Urban
the completion of its preliminary design. However, only Renaissance B03) from High-Tech Urban Development
limited project information available in the conceptual Program funded by the Ministry of Land, Transport &
phase was used for the test, i.e., site area, building area, Maritime Affairs of the Korean Government.
gross floor area, building-to-land ratio, floor space index,
number of building, stories, uses, etc. Test result was REFERENCES
satisfactory considering that the common accuracy of [1] Lam, K.C., Wang, D., and Lam, M.C.K., “The Capital
conceptual estimate is between +75% and -25% [8]. The Budgeting Evaluation Practices (2004) of Building
construction cost estimate showed a difference of -6.5% Contractors in Hong Kong”, International Journal of
compared with the detailed estimate. Regarding the total Project Management, Vol. 25(8), pp. 824-834, 2007.
capital costs and rate of return, small differences of +3.5% [2] Hendrickson, C., Project Management for Construction
and -3.4% were found, respectively, compared with the – Fundamental Concepts for Owners, Engineers,
feasibility study, assuming same revenues. The test took Architects and Builders, Carnegie Mellon University,
only two and a half hours, which is very significant time Pittsburgh, PA, 1998.
saving. The system performance may be greater due to its [3] Gunduz, M., Ugur, L.O., and Ozturk, E., “Parametric
revision control features. Cost Estimation System for Light Rail Transit and Metro
Trackworks,” Export Systems with Applications, Vol. 38 (3),
6. CONCLUSIONS pp. 2873-2877, 2011.
The current study developed a comprehensive but practical [4] Flood, I. and Issa, R.R.A., “Empirical Modeling
economic evaluation system for capital building Methodologies for Construction, Journal of Construction
construction projects, which can be used in the conceptual Engineering and Management, Vol. 136(1), pp. 36-48,
phase. This system has unique aspects: 1) the efficient 2010.
management of information flow based on the integration [5] Sonmez, R., “Conceptual Cost Estimation of Building
of analytical steps; 2) the combined cost estimating method, Projects with Regression Analysis and Neural Networks,”
which enhances the confidence in the final estimate; 3) the Canadian Journal of Civil Engineering, Vol. 31(4), pp.
automated tabulation and scheduling of cash flow; 4) trade- 677–683, 2004.
off between automation and flexibility; and 5) the revision [6] Kiziltas, S., Akinci, B., and Gonzalez, C., “Comparison
features. Regarding the estimating method, the comparison of Experienced and Novice Cost Estimator Behaviors in
of the two types of estimates was so favored by Information Pull and Push Method,” Canadian Journal of
practitioners during the case study. There are Civil Engineering, Vol. 37(2), pp. 290-301, 2010.
improvements to make for future study: the generic S- [7] Karshenas, S., “Predesign Cost Estimating Method for
curve needs to be developed for different types of building Multi-story Buildings, Journal of Construction Engineering
and more historical data and assemblies need to be and Management, Vol. 110(1), pp. 79-86, 1984.
accumulated, which improve the accuracy of the estimates. [8] Taylor, J.C., Project Scheduling and Cost Control:
Planning, Monitoring and Controlling the Baseline, J. Ross
Publishing, Fort Lauderdale, FL, 2008.

472