Forecasting Construction Duration
Construction duration can be defined as one or a combination of the following:
1) The construction duration arising from critical path in which duration for items of work or activity in sequence cannot be reduced further (Barrie and Paulson, 1992).
2) Duration means the time required to complete a specified task or activity. And, construction duration is the time determined by the owner’s needs to occupy, utilize, or rent the completed space of the project (Callahan et al., 1992).
3) Construction duration is a duration resulting from an examination of one or more methods of carrying out the works on the basis of minimum cost, it is usually estimated in the first instance for normal condition (Pilcher, 1992).
4) Construction duration refers to a given time to execute and complete item(s) of work using all project information and resources within an estimated or predicted cost (Kwaku, 1994).
5) Construction time can be defined as the elapsed period from the commencement of site works to the completion time of building to the client. It is usually specified prior to the commencement of construction (Nkado, 1995).
In this research, construction duration is defined as the time frame given by the owner for the contractor to complete the project under normal work conditions, normal practice of construction, and based on the minimum costs. It starts when the contractor receives the instruction to proceed and ends at the completion of construction works on site. It also includes delays caused by unanticipated circumstances, e.g. alteration of works (changed conditions and change orders), extra works, supply of materials, location, weather, and site work conditions. Major changes that after the scope of work significantly are not included.
Scheduling and Schedulers
Control of construction duration needs a clear systematic plan and commitment on the part of the people involved (McNulty, 1982). The systematic plan is known as schedule. The scheduling is the determination of the timing of activities and follows logically from the planning process (Callahan et al., 1992). It is normally used for controlling construction duration (Callahan et al., 1992). To both the owner and contractor, scheduling plays an important role in financial proposal and budgeting (Peurifoy and Ledbetter, 1985; Kaka and Price, 1991). The schedule is prepared by the scheduler and/or planner. In preparing a schedule, the scheduler or planner may meet or discuss with some people for crucial information, e.g. estimator, manager, superintendent, sub-contractor, architect, engineer, owner, and materials’ suppliers (Callahan et al., 1992). They may need to study the contract, drawings, specification, and perhaps, conduct site reconnaissance. Further, they need to know about manpower and productivity. (Callahan et al., 1992; Pilcher, 1992). Apart from the schedule, it is also necessary to prepare systematic monitoring to provide early warning of restraints as well as imaginative action to overcome them (McNulty, 1982).
Nowadays, computers may assist the scheduler or planner by storing and sorting the information, as well as performing mathematical calculations but they do not provide the intellectual direction (Callahan et al., 1992). Human skill and experience are still necessary, i.e. the scheduler or planner may use experience of similar projects in estimating the construction duration (Pilcher, 1992). The scheduler or planner may use their judgement according to various constraints, e.g. location and access, weather, space and site work conditions, complexity of the project, quality of workmanship, delivery of materials, and economic or marketing conditions (Pilcher, 1992).
Forecasting Construction Duration
The basic inputs for project scheduling are:
4) resources (materials, manpower and productivity); and
5) other constraints, e.g. site conditions and weather (Burgess and White, 1979; Fisk, 1982; McNulty, 1982; Peurifoy and Ledbetter, 1985; Ashworth, 1988; Barrie and Paulson, 1992; Callahan et al., 1992; Pilcher, 1992).
The contract may clearly specify the completion date. Therefore, the scheduler or planner may use this time frame to prepare the schedule together with other factor constraints. First, it is usual to prepare the schedule for normal conditions by assuming one or more methods of carrying out the works on the basis of minimum cost (Nkado, 1992; Pilcher, 1992). When it is necessary to shorten the duration for an activity or a project, crashing may be done, e.g. increase manpower and overtime working. This process leads to increase in construction cost (Puerifoy and Ledbetter, 1985; Barrie and Paulson, 1992; Pilcher, 1992).
Drawings contain physical features of the project: 1) function; 2) height; 3) systems (e.g. plumbing, fire fighting, and lighting); and complexity (Ireland, 1985; Ashworth, 1988; Pilcher, 1992; Nkado, 1992). In other words, the information on the drawings is presented in form of dimensions and descriptions (Ashworth, 1988). Materials and installed equipment for the project, as well as the plant and construction equipment can also be known from the drawings (Ashworth, 1988; Peurifoy and Ledbetter, 1992). Meanwhile, the contract may specify the constraints of work, e.g. construction cost, duration, payment, inspection, method or conditions of works, delays, and damages. In addition, the specification may address the quality of materials, workmanship, and method of working (Barrie and Paulson, 1992).
2. Resource scheduling
The scheduler or planner has to allocate the resources, i.e. materials and manpower (or working team) to each activity or task at the proper time they are needed. Thus, the resource schedule deals with levelling and allocation of all necessary resources (Peurifoy and Ledbetter, 1985). The former smoothes out the peaks and valley in resource use within the project duration. The latter determines the shortest project duration consistent with the limited resources.
4. Other judgements
They also have to provide timing for preliminary works, e.g. construction plant, and mobilization (Peurifoy and Ledbetter, 1985). This often includes time for remedial works and site clearing after completing the construction.
Site reconnaissance enables the scheduler or planner to adjust the schedule against various constraints, e.g. location and access, weather, space and site work conditions, complexity of the project, quality of workmanship, delivery of materials, and economic or marketing conditions (Barrie and Paulson, 1992; Pilcher, 1992).
Some Methods for Scheduling
There are many forms of schedules, e.g. Gantt or bar chart, Critical Path Method (CPM), and Program Evaluation and Review Technique (PERT). Each has its own advantages, disadvantages, and application areas for which it is most appropriate. They are means of visual presentation of a construction program (Pilcher, 1992), and used for project planning, management, and control (Burgess and White, 1979; Fisk, 1982; McNulty, 1982; Barrie and Pualson, 1992).
Gantt chart simply represents the activity and its duration by a bar. It is sometimes called Bar chart. Gantt or bar chart can include a great deal of anticipated and actual information for; 1) cashflow; 2) manpower and/or manpower by trade; and 3) productivity (Barrie and Paulson, 1992; Callahan et al., 1992).
In CPM scheduling, a construction project is sub-divided into several activities. An activity is a single work step that has a recognizable beginning and finishing or ending (Callahan et al., 1992). In other words, the activity is a time-consuming task. The basis of CPM is network diagram, i.e. which needs nodes and arrows. It deals with four aspects: 1) activities identification; 2) logical sequence,; 3) network construction; and 4) allocation of resources (Barrie and Paulson, 1992). Callahan et al. (1992) divided the development of CPM schedule into six phases:
1) understanding the project;
2) conceptual approach definition;
3) physical creation of the schedule;
5) refinement; and
Further, physical creation of CPM scheduling is divided into eight steps: 1) select software;
2) divide project into several activities and sub-networks;
3) develop responsibility code;
4) develop information code;
5) develop specific sub-networks;
6) build or plot the logic diagram;
7) numbering the activities; and
8) linking the sub-networks together.
The longest interconnected path of activities through the network called “critical path” becomes the project duration. On the critical path means that the duration for items of work or activity on the sequence cannot be reduced further without paying extra costs (Peurifoy and Ledbetter, 1985; Barrie and Paulson, 1992).
PERT is like CPM scheduling. It uses logical diagram to analyze performance time. It overcomes difficulties associated with duration of activities which could not be estimated reliably. PERT enables the scheduler to estimate the most probable project duration and the probability that the project or any portion of the project will complete at particular time. PERT focuses on events or nodes, called event-oriented. Unlike the CPM, it requires three estimates of duration for each activity: 1) optimistic (high productivity); 2) pessimistic (low productivity); and 3) most likely duration (Peurifoy and Ledbetter, 1985; Barrie and Paulson, 1992; Callahan et al., 1992; Pilcher, 1992).
Factor Affecting Construction Duration
The following are some factors affecting construction duration and its estimate.
1. Size of project
Size of the project can be represented in terms of functional or floor area, i.e. in ft2, or m2. The larger the building size, the more complex the construction, thus needing longer duration to complete (Sadashiv, 1979; Ireland, 1985; Ashworth, 1988; Pilcher, 1992; Nkado, 1992).
Function implies type of building and required engineering systems, e.g. plumbing, fire fighting, and lighting (Ashworth, 1988; Pilcher, 1992). It is an important facet in designing of construction project (Ashwoth, 1988). Function of a building implies business target that the building serves. It can be considered as qualitative variables, e.g. office, retail, and other buildings (Nkado, 1992).
Height of building, represented by number of floors (or storeys) affects the construction duration (Sadashiv, 1979; Ireland, 1985). The height of building indicates construction technique, major equipment used, and construction sequence (Sadashiv, 1979; Callahan et al., 1992).
Complexity implies unfamiliarity with work (Pilcher, 1992). The complexity of building impacts the form of construction, i.e. building frame, foundation, and systems (Ireland, 1985; Ashworth, 1988). Complexity can be represented in form of construction equipment, method and sequence (Sadashiv, 1979; Callahan et al., 1992; Chan and Kumaraswamy, 1995).
Quality can be classified by variables or attributes, i.e. appearance, strength, stability, materials used, performance finish. Appearance of the building, e.g. external facing is one aspect of quality (Ashworth, 1988). Sadashiv (1979) considered number of major finishing works in duration forecasting instead of a defined quality index.
The location of the building has a significant effect on the construction duration (Chan and Kumaraswamy (1995). It reflects restrictions or easements that exist and availability of services (Burgess and White, 1979). It effects supply of resources, e.g. materials, and equipment (Sadashiv, 1979). Consequently, it also effects the use of major equipment (Sadashiv, 1979), and productivity on site (Callahan et al., 1992).
There are other possible factors affecting the construction duration, e.g. construction planning (Sadashiv, 1979; Ireland, 1985), design-construction interface coordination (Ireland, 1985), dispute per unit of time (Ireland, 1985). Type and/or variation to the contract refers to risk allocation management structure and payment modalities (Burgess and White, 1979; Ireland, 1985; Chan and Kumaraswamy, 1995). By contrast, Walker (1994) concluded that client related factors have more significant affect on speed of construction, or construction duration, than the contract type. Callahan et al. (1992) pointed out that quality of supervision, labour training and motivation, can also be affecting factors. Al Tabtabai et al. (1997) developed the models for expert judgment in forecasting construction project completion. The factors are: 1) performance of management; 2) cash flow situation; 3) material and equipment availability; 4) labour productivity; 5) weather and environment influences; 6) rework, extra work, and work difficulty; 7) percentage of work completed; and 8) trend in schedule variance.
Delays and their Causes
Project duration is normally specified by the owner. The completion and operation of many buildings are restricted as they are seasonal in nature, e.g. a school must open in September, a retail outlet must open for booking in August to meet the Christmas sale season, and an apartment must open in May to reach the spring market (McNulty, 1982). Construction duration affects the financial interest to the owner, e.g. selling price, and on-site management. Then, the contractor is traditionally responsible for the detailed planning and scheduling to ensure the completion of the project within the owner’s time frame. However, the actual construction duration consist of two parts:
1) contract time; and
Delay is the time during which some parts of construction project has been extended or not performed because of unanticipated circumstances (Barrie and Paulson, 1992). When necessary, the contractor may need to allocate an additional budget for corrective actions to maintain the schedule, otherwise, it may cause liquidated damages charges against the contractor for failure to meet the owner’s requirements (Fisk, 1982; McNulty, 1982).
Alteration of working drawing is one of a major factor affecting the construction duration, i.e. it may cause a delay beyond the contract time (Sadashiv, 1979; Chan and Kumaraswamy 1995). Barrie and Paulson (1992) summarized causes of delay into four areas:
1) changed conditions and change orders;
2) extra works;
3) owner or his/her agent; and
4) third party contractors.
However, other possible delays may result from location, weather, site work conditions, labour productivity, sub-contractor, supply and delivery of materials (Burgess and White, 1979; Sadashiv, 1979; Callahan et al., 1992; Chan and Kumaraswamy, 1995).
Relationship between Cost and Time
There is a relationship between cost and duration (or time). In construction planning and scheduling, alternative duration and costs for activities are always considered. For example, when it is necessary to shorten the duration of activities or project, the cost may increase. This is called “crashing“.
Sadashiv (1979) found that some of the independent variables used in regression model for cost forecasting also have major affects on construction duration, e.g. height, and types of major equipment. However, he used the number of major finishing works in his duration forecasting model instead of the quality index as used in the cost forecasting model.
Ireland (1985) found a relationship between construction time and cost. In his study, the construction time and cost are dependent on some common independent variables:
1) construction planning during design;
2) variation to the contract; and
3) complexity of form of construction.
In addition, number of storeys, design-construction interface co-ordination, and dispute per unit of time also affect the construction time while architectural quality, and use of nominated sub-contractors affect only the construction cost.
Kaka and Price (1991) found strong relationship between the cost and duration of construction projects that can be used in contractor’s budgeting systems and corporate financial model. Investors of project can utilize the relationship for financial appraisal and budgeting. Estimating of project cost can be used to derive the expected duration and vice versa.
Pre-design Estimating of Project Duration
Pre-design estimating of construction duration is important. Chan and Kumaraswamy (1995) noted that pre-contract determination of the construction duration is essential for proper cash flow forecasting by both the contractor and the client. It can facilitate optimal resource allocation, financial planning, profitability and efficiency of capital flow within a pre-determined time limit.
Based on the inputs required for scheduling, all the current methods of scheduling seem to be efficiently applied only when the detailed design is completed. Normally, the contractor must complete this planning prior to bidding for the project (Peurifoy and Ledbetter, 1985). Without sufficient information, the schedule can be prepared based on only the best guess, i.e. using experience of similar projects in estimating the construction duration (Pilcher, 1992).
Nkado (1992) established a computerized construction time information system for planning of buildings at the early stage of design. Two key assumptions are:
1) the building team is competent and efficacious in setting up the building process and working within local norms and organization form to bring the project to a successful completion; and
2) the frame of reference for construction times is based on the overall time consistent with the minimum direct cost of construction to the contractor.
Al Tabtabai et al. (1997) developed the multi-regression and neural network models to capture the decision-making procedure of project experts involved in schedule monitoring and prediction. The models were applied to a multi-storey building under construction. They provided the convenient and realistic generation of revised schedules at appropriate junctures during the progress of the project. The accuracy of the models mainly depends on the soundness of the underlying expert decision, i.e. inputs generated from judgment of the experts are not free of bias. However, they concluded that modelling construction experience to use in future projects can help significantly in achieving project objectives.
A model for forecasting the construction duration should be valid in application for a reasonably long period of time without the effects from changes in price level. The price indices may not be applied or included in the model. The model should be slightly affected by change in construction technology in long term application. Today, the construction industry is still labour intensive. Automation which has reduced the labour force in manufacturing industry, has not been matched in the construction industry (Ashworth, 1988). This means
construction technology has changed only slightly over time.
A variety of methods and techniques for construction planning and scheduling exist but they are based mainly on the completed design and details of project, e.g. Gantt chart, CPM, and PERT. Construction schedule is normally prepared by the contractors at the time they submit bids. Without sufficient information, the scheduling and forecasting of construction duration is based on the experience of the planner or scheduler. A number of researches found strong relationship between construction duration and cost. This leads to the possibility to build a model for forecasting the construction duration at pre-design phase. The model may consist of part of variables as they are used in the cost forecasting models. The main variables shall consist of building features, e.g. function, structural system, height, foundation, exterior and interior finishing. Adjustment of the construction duration by means of indices is not necessary.