Change management in construction projects using Building Information Modeling

¶ … Change Management in Construction Projects Based on Building Information Modeling

Valeh Moayeri, Osama Moselhi, and Zhenhua Zhu

Department of Building, Civil, and Environmental Engineering, Concordia University, Montreal, Canada

Design changes after awarding the contract are typical during the construction phase of a project. The changes do not only affect the original project schedule, but also produce additional construction costs. They cause the potential contractual disputes among project parties (e.g. general contractor, sub-contractor, designers, and owners). As a result, the design changes have been considered as one major cause of leading to the project failure. In order to effectively manage design changes, it is necessary provide the prompt feedbacks about the impacts of the changes on the project schedule and cost upon receiving the design change request. The objective of this paper is to propose the use of building information modeling (BIM) technologies to facilitate design change management. Specifically, any potential design change in a construction project is virtually implemented into its building information model. Then, the components affected by the change are identified in the model. These components are further linked with construction cost and schedule information. This way, the parties in the project has a preliminary idea about the impacts of the design changes and their ripple effects, which helps the owner to make a final, well-informed decision on approving or rejecting the requested changes to the project\'s design. The research work presented in this paper has been tested in the case study of a 3-story residential building project. The results showed that the impacts of the changes in terms of time and cost could be reasonably estimated before implementing the changes in the project.

INTRODUCTOIN

Design changes, including any addition, deletion and modification in the design or construction of a project after awarding the contract, are expected during the construction phase of a project (Yap et al. 2015).. This is partly because project owners might change their minds in line with the changing economic climate to meet the market demands (Ogunlana et al. 1996). In addition, designers might produce inadequate or inconsistent detailing of drawings, which lead to a large margin of error and omission and create problems of coordination between the architectural, structural, mechanical, electrical and other systems (Hegazy et al. 2001). For example, the fire-rated walls have to be relocated due to and architectural restrictions or the air supply ducts need to be rerouted under the limitations in available space (Pilehchian, Staub-French, 2012).

The construction project performance in terms of schedule and cost is significantly impacted by design changes. In fact, they have been considered as one of the key reasons for project schedule delays and cost overruns (Jackson, 2002). For example, Chang (2002) once reported that cost increased on average of 24.8% and schedule increased on an average of 69% as a result of design changes after surveying four sampled projects in California. Similarly, in process plant and offshore oil and gas projects, Long (2016) noted that design changes might result in the need of more construction labor man-hours to complete the project and the increased labor requirements could then lead to addition loss of productivity as a result of crowding, trade stacking, dilution of supervision, etc. Yap et al. (2015) concluded that design changes are on-going problems that continue to raise concerns in the construction industry.

In order to manage design changes in an effective manner, it is necessary to give prompt feedback about the impacts of the changes and their related ripple effects on the project performance. However, this might be difficult, especially considering the fact that the owner\'s requests for design changes are usually made at short notice (Olawale and Sun, 2010). So far, most of existing research studies have been conducted for change management. They focused on creating the guidance for best practice in change management (Moghaddam, 2012; Erdogan et al. 2005; CII, 1994) and quantifying the overall impacts that changes have on construction project performance (Serag et al. 2010; Ibbs, 2012; Ibbs et al. 2007). Recently, the idea of developing project information models has been proposed. The information modes are expected to facilitate the coordination of design information through the management of design changes (Xue et al. 2011; Hegazy et al. 2001).

The main objective of this paper is to investigate the impacts of design changes and their ripple effects on the project performance in terms of cost and schedule with the aid of building information modeling (BIM) technologies. Specifically, suppose the project building information model has been created in advance. Any potential design change requested by the owner or designer in the project could be virtually implemented into its building information model. Then, the new model (i.e. as-changed model in this study) is compared with the original model (i.e. as-planned model) to identify the components directly and indirectly affected by the change. These components are further linked with construction cost and schedule information, which provides a preliminary idea about the impacts of the design changes and their ripple effects on the project in terms of schedule and cost. This information could help the project owner to make a final, well-informed decision on approving or rejecting the requested changes to the project\'s design and avoid potential construction disputes between different parties. The research work presented in this paper has been tested in the case study of a 3-story residential building project. The results in the case study showed that the impacts of the changes in terms of time and cost could be reasonably estimated with the BIM technologies.

RELATED WORK

A change has the impacts on the project delay and cost overrun (Arain and Pheng, 2005). Also, it affects the quality of construction work, labor productivity, etc., which cause legal disputes between different parties involved in the project (Mirchekarlou, 2012). Moreover, one change itself could be further spread and escalated. As a result, a series of additional changes are produced as ripple effects. However, such ripple effects of the change are not always easy to be identified. It requests enough experience and expertise to predict or estimate the impacts of all possible consequences of the change in the project.

So far, several research studies have been proposed for the purpose of quantifying the impacts of the change in the project. They relied on different computing techniques, such as regression/statistical analysis, artificial intelligence, system dynamics, decision tree, and case-based reasoning. For example, Hanna et al. (2002a) grouped the factors that correlate with whether a project is impacted by change orders through statistical hypothesis testing, and then developed models to predict the probability of a project being impacted using logic regression techniques. In one of their other studies, they also presented a hybrid approach to quantify the impact of change orders on construction projects using statistical regression and fuzzy logic (Hanna et al. 2002b). Lee et al. (2004) developed a decision tree model to classify and quantify the labor productivity losses due to the cumulative impact of changes for electrical and mechanical projects. Similarly, Moselhi et al. (2005) proposed the idea of evaluating the percentage loss of labor productivity due to changes with an artifical neural network. Yitmen et al. (2006) presented a prototype expert system to quantitatively model how different changes affect the time and cost of a project in North Cyprus construction industry. Motawa et al. (2007) integrated a fuzzy logic-based change prediction model with the system dynamics to evaluate the negative impacts of changes on construction performance and also discover the \"cause-and-effect relationships\" of the change events. Isaac and Navon (2008) created a change control tool to identify the scope of the implications of a change as soon as it is proposed. The feasibility of the tool in practice has been illustrated in several pilot studies (Isaac and Navon, 2008). Arian (2008) presented a knowledge-based decision support system for the management of changes in educational building projects in Singapore. The system was expected to assist project managers by providing accurate and timely information for their analysis and decision-making of change orders (Arian, 2008). Serag et al. (2010) created two regression models; and the models were used to estimate the impact of change on a contract price in heavy construction projects in Florida. Zhao et al. (2010) developed a change prediction model to help project management teams to manage project changes in a proactive and efficient manner. In the model, they relied on the activity-based dependency structure matrix (DSM) to model the process that may occur as a result changes, and then Monte Carlo Simulation to analyze the change probability of activities involved in construction projects (Zhao et al. 2010). Nahod and Radujkovic (2011) developed a Dynamic Planning and Control Methodology (DPM) to facilitate the objective decision-making in approving changes in construction practice. Moreover, Ibbs (2012) studied hundreds of project data collected from the work of Leonard (1988) and Ibbs (2005), and produced a set of curves and reference points that indicate the relationships between the amount and likelihood of change and the amount and nature of its impacts.

Recently, the research studies of using BIM technologies to quantify the impact of a change on the project performance has been initiated. Building information model is a digital representation of building physical and functional characteristics, which is composed of digital objects corresponding to real world components such as doors, walls, and windows with associated relationships, attributes and properties (NBIMS, 2007; IBC, 2011). There are several benefits regarding the use of, including the better management of design documents and ability to simulate construction operations prior to physical implementation. Therefore, BIM is expected to help to trace the pattern in the changes that happen during the life cycle of a project and identify the types of changes (Akcamete et al. 2009; Koch and Firmenich, 2011). Langroodi and Staub-French (2012) presented a case study, which examined change management in the context of a multi-disciplinary collaborative BIM environment during the design and construction of a fast-track project. They found that it was possible to document numerous changes and identify their essential attributes with BIM (Langroodi and Staub-French, 2012). Based on these findings, they further developed an approach to represent, coordinate, and track changes within a collaborative multi-disciplinary BIM environment (Pilehchian et al. 2015). In addition, Liu et al. (2014) developed an integrated framework for embedding change management in BIM.

OBJECTIVE AND PROPOSED METHODOLOGY

Figure

1

: Composition of a Model Componentthe objective of this research study is to investigate the potential of using the BIM technologies to identify, analyze, and visualize the impacts of a change and its ripple effects. The basic idea is that the change requested by the owner or designer is first virtually implemented in the building information model. Then, the new model (i.e. as-changed model in this study) is compared with the original model (i.e. as-planned model). The comparison is conducted at the model component level. Therefore, the project work breakdown structure (WBS) is required to be defined at a micro level, where each component in the model has its own ID, location, time, and cost information, as shown in Figure 1.

When a change has been requested in the as-changed model, the components directly affected by the change are identified through a simple component-by-component comparison. Specifically, both as-planned and as-changed models are checked. If the components are found in the as-planned model but not in the as-changed model, it means these components are removed by the change. If the components are found in the as-changed model but not in the as-planned model, these components are newly added by the change. If the components exist in both models, their specifications are further checked to find whether they are modified or not by the change. As a result, the direct impacts from the change are determined.

In addition to the direct impacts from the change, its ripple effects are also identified by finding the components indirectly affected by the change. In doing so, all the components that are connected to those directly affected by the change are detected. If the modifications are found on those components, they are recorded. Moreover, further detections are made on those components, until there are no more modifications that have been found in their connections. As a result, the list of all the components indirectly affected by the change could be produced. The design change path or the change sequence is determined. For each component due to the change, it can be classified into three categories: addition, deletion, and modification.

The way in which this occurs is thus: the proposed model traces indirect impacted components (via the ripple effect) by first finding the level (1st floor) at which the selected direct change is located. The model then checks the components that are connected to the direct change in other levels, searching level by level until every floor is scanned. The model starts at the first level and finds connected components to the selected direct change, identifying all found components and matching component IDs with the component IDs in the impacted component database. Then it selects all the impacted components one by one and identifies the connected components in the next level (2nd floor). This process is continued up the levels. From the components found in the next level, the model identifies the impacted components using the same matching process. Then it selects the first impacted component in the 2nd level and finds the components that are connected to it in the next level (3rd floor). The process is the same to the last level. When the proposed model gets to the last component in the last level it saves this path. Then it will follow the saved path to return to the first impacted component in the first level. There it selects the next impacted component in the first level and performs the same procedure to get to the last level and again get back to the first level by using the second saved path. The model again checks for the next impacted component till there are no more impacted components left in the first level that the selected direct change is connected to.

To summarize, the proposed model for each select individual direct change traces all the components in the horizontal level and in then in the vertical level. It starts with one of the impacted components in the first horizontal level and it goes horizontally down levels till there are no more levels left and it finds the path of impacted component throughout all the levels; then it goes backward to get to the first level. Then, the same process is enacted for the all impacted components in the first horizontal level till there are no more impacted components in that level left. The ripple effect works as a tree diagram: on top of the tree the individual direct change is located, and in the other levels are all the other components that are connected to the direct change.

Based on the list of the components affected by the design change, the time and cost impacts on the project due to the change are quantified. The time impact analysis is to calculate the number of days that need to be added or reduced from the project\'s total duration. When a component is added or modified, the new duration for the component is estimated and inserted into the project schedule. When a component is deleted, its duration information is removed from the project schedule. Therefore, the project schedule is updated correspondingly.

Also, the costs incurred due to the change is estimated. They includes the direct cost, indirect cost, and impact cost, considering the change affects construction productivity on the components even if they are not modified in the project. The direct cost is composed of those from project labor, material, equipment, and sub-contractors. The indirect cost mainly includes project overhead. The impact cost here is focused on productivity and time-related ones that represent the loss of productivity on percentage change in the project. All the costs are added up to provide the basic project cost information about the change. It allows the project owner to compare the cost of the change with the project original cost to see if the requested change is beneficial or worth.

CASE STUDIES

The proposed methodology was developed as add-ons in the environment of Autodesk Revit 2014. The add-ons were implemented using the Revit Application Programming Interface (API). They could automatically 1) detect the changes between the as-planned and as-changed models; 2) detect the cause and effect relationship between the changed components in the model; 4) calculate and quantify the ripple effects in terms of project cost and schedule; and 5) filter the change impact reports.