Crusell Bridge: BIM Implementation in the Construction Phase

1. Project Introduction

The Crusell Bridge is a cable-stayed bridge managed by Helsinki’s public works department, connecting the west side of Jatkasaari to the city of Ruoholahti. Jatkasaari, a part of the former West Harbor near central Helsinki, is evolving into a new coastal district. To facilitate the development of a residential complex with 9,000 units, freight operations were relocated elsewhere, necessitating a new highway bridge.

Figure 1: Rendering of the cable-stayed bridge at Helsinki Harbor

Construction of the Crusell Bridge began in autumn 2008 and was scheduled for completion by the end of 2010. The design was executed by WSP, a Finnish company, while Skanska Civil handled the construction. The bridge features two asymmetric spans measuring 92.0 meters and 51.5 meters respectively, totaling 143.5 meters in length, with a net width of 24.8 meters. The superstructure consists of longitudinal prestressed concrete beams, and the transverse structure integrates composite steel and concrete, as depicted in Figures 2 and 3.

Figure 2: Rendering of the bridge at night

Figure 3: Architectural model of bridge structure

Throughout design and construction, the project team utilized Building Information Modeling (BIM) technology alongside lean construction principles and tools. This case study focuses on the construction phase, highlighting two main aspects:

• Extensive use of building information models for steel beam and reinforced concrete fabrication, supply chain monitoring and management for assembly components, template and temporary support structure design, quality control via laser scanning, and 4D animation simulations for construction planning.
• BIM’s support for lean construction methods, including site production management and implementation of the Last Planner System (LPS).

Table 1: Crusell Bridge project team and basic information

The design competition for the Crusell Bridge was announced by the city of Helsinki in winter 2001, aiming to find a high-quality bridge solution that reflects local character and respects the landscape. Although a British design firm won the competition, the project was awarded to the Finnish company WSP, which placed second. Design development halted at 60% completion in late 2004 due to financial constraints. After a four-year suspension, the Public Works Department appointed its Construction Management Department in 2008 to issue a tender for a general contractor; Skanska Civil was selected.

Since only 60% of the design documents were ready, the detailed design had to progress simultaneously with construction, which began in autumn 2008. The project was expected to finish by September 2010.

Figure 4: Overall project timeline

The contract followed a Design-Bid-Build (DBB) model, surprisingly despite the incomplete design documents. This approach allowed early selection of manufacturers, enabling them to influence final design details. For example, Ruukki, the steel structure manufacturer, contributed its expertise to complete detailed steel drawings. The success of the steel components during construction validated this strategy.

2. The Crusell Bridge Project as a Learning Experience

The Crusell Bridge project served as a BIM learning platform for all involved, testing many novel solutions and technologies. Initially, the designer created some bridge models to support the visual design competition. Due to the conceptual design’s heavy use of steel structures and precision requirements, the owner was encouraged to adopt modeling for better results. This led to modeling not only steel components but all construction elements, including reinforced concrete structures onsite.

Consequently, the project became an experimental endeavor for both the owner and designer. It was the owner’s first bridge project fully utilizing BIM, including time and management aspects. Previous models were simpler reinforced concrete bridges, while the Crusell Bridge’s bending geometry, diagonal tension cables, and composite steel and concrete structures marked a significant increase in complexity.

Bridge engineering differs considerably from industrial and residential construction due to more complicated structures. While computer modeling for structural analysis is common, BIM use in bridge engineering was less frequent at the time. Few BIM applications could accurately simulate the complex geometry of modern bridges. Although some bridge modeling software existed, integrating design with BIM was a new concept.

Initially, the contractor was aware of the model but did not receive it from the designer. During bidding, the designer provided a simplified network model representing basic bridge components. Once Skanska was hired, they received the full model created with Tekla Structures and committed to maximizing its use during construction, including 4D planning and temporary structure modeling. While experienced with modeling in residential and industrial projects, this was their first application in bridge engineering. Other project parties, such as subcontractors, surveyors, and suppliers, also found model use a new experience.

Skanska reported positive outcomes and considered this a valuable learning experience. The pioneering BIM application attracted support from Tekla, the major BIM software supplier, which provided assistance throughout design and construction. Tekla helped team members learn new software features including synchronous network model sharing, 4D planning, integration with suppliers’ factory management software, and direct output of manufacturing data to computer-controlled equipment.

Overall, the project became a unique learning process involving all participants, fostering the adoption of new workflows and accumulating valuable experience.

3. Data Exchange Capabilities

Table 2: BIM and other software used at various stages

BIM software was introduced at the start of the second design development phase in 2008. Prior to the 2004 design suspension, software tools were treated as independent. Visualizations were produced using 3D Studio Max, which lacked BIM’s parametric capabilities. After resuming, the team adopted Tekla Structures, recognizing BIM’s maturity for complex bridge projects.

Before 2004, data exchange was not a major concern. However, with new project partners joining in 2008, interoperability became critical. A driver program was developed to enhance software exchange functionality, allowing broader usage across partners. As a result, the owner, designer, general contractor, and major subcontractors all adopted Tekla Structures, reducing synchronization issues.

Data interchange also occurred with other software such as Trimble RealWorks, Vico Control, PERICAD, Enforcement List v3.1, and manufacturing ERP systems. Geometry-only exchanges, like between Trimble RealWorks and Tekla, used DWG files. For richer information exchange, such as between Tekla and PERICAD, IFC files were utilized. When textual or numerical data was sufficient, or geometry was parameterized (e.g., steel bar shapes), simple ASCII files were generated from Tekla models. Figure 5 illustrates these exchange formats.

Figure 5: Information exchange through data translation

4. Model Synchronization

Large engineering projects involve multiple participants, each developing specific aspects of the building information model. To improve communication and data exchange, Tekla introduced a model synchronization feature using a central server. The Crusell Bridge was the first bridge project to apply this feature. As seen in Figure 4, synchronization was crucial since detailed design and construction occurred simultaneously, which is rare in traditional DBB projects.

Synchronization between Skanska’s construction model and Ruukki’s manufacturing model was vital, reflecting the common need for manufacturers to develop detailed manufacturing drawings. The client began using synchronized servers in autumn 2009.

Synchronization typically occurred weekly but was not mandatory. When designers made significant design changes, they notified the construction team, who could immediately update their models with the latest revisions. Contractors maintained models of temporary structures, templates, and other items; only modified components were updated. Project information personnel filtered changes, assessed their impact, and often received updates before official client approval, enabling proactive preparation.

Synchronization between construction site and subcontractor models also happened, though less regularly, primarily to accommodate changes. Figure 6 illustrates information exchange via synchronization.

Figure 6: Information exchange through synchronization

Successful synchronization required not only technical solutions but also management agreements covering editing rights, objects, and timing. The agreed workflow included:

1. WSP (designer) uploading model changes to the synchronization server.
2. Skanska uploading progress changes.
3. Ruukki uploading manufacturing updates and progress data (order, production, delivery dates).
4. All participants downloading the changes and syncing them into their models.

Some issues arose during the project. For example, when Skanska upgraded to Tekla Structures version 15, synchronization of steel reinforcement data conflicted with WSP’s version 13 model. This limited synchronization to one-way updates until resolved.

5. Using BIM During the Construction Phase

This section explores various methods of applying BIM for management and organization during construction, serving as information resources and supporting lean construction practices.

Although not a large-scale project, the Crusell Bridge’s design complexity allowed BIM to fulfill important engineering roles. All bridge structures, including every steel bar and temporary support, were modeled. Skanska maintained the model on a server at the construction site office, with a civil engineer serving as the “contractor information personnel” responsible for disseminating information and updating the model.

Skanska did not manage model maintenance directly onsite because the design was incomplete at project start, preventing comprehensive planning. Initially, the construction team hesitated to embrace the model, uncertain of its benefits or how to use it. However, as each object was modeled, the contractor’s model evolved alongside the designer’s, becoming the primary information source for construction teams, covering dimensions, visualization, procedures, and material logistics.

Though the initial model was created by the designer, Skanska also refined it to reflect construction progress. Models were used extensively for task management, sequencing, and visualization. All temporary structures — support towers, piles, templates, equipment — were modeled. For example, carpenters referred to the model to understand complex polygonal geometries with hyperbolic curves before forming concrete templates.

Due to limitations in Tekla Structures at the time, complex hyperbolic shapes were generated in 3Ds Max and imported via DWG files as geometric references. This issue was resolved in newer Tekla versions, a direct result of the company’s involvement in the project. Tekla viewed the Crusell Bridge as a testing ground for software construction management functions.

Visualization

Using building models for visualization is the most common and immediately beneficial application. The 3D model helped all participants understand design concepts and details faster and more thoroughly than traditional drawings. Construction workers could access the model as an extension of their work, occasionally visiting offices to review detailed information such as template positioning, anchor cables, and reinforcement locations.

Figure 7: Detailed cross-section showing the relationship between steel bars and cast hardware, such as large anchor cable components

Design and Planning of Temporary Structures and Collision Detection

Initially, template drawings provided to the site lacked extended support towers and temporary structures like scaffolding. To address this, the construction team modeled all missing temporary structures directly into the design model maintained onsite, including template supports and tower crane tracks. This approach simplified interpretation, identified many conflicts, ensured accurate quantities, integrated tasks into the schedule, and enabled 4D sequencing visualization.

Collision checks were not only performed at design completion but also continuously during construction by incorporating temporary structures and templates. This prevented numerous potential conflicts on the bridge, saving costs and avoiding delays.

For example, template supplier PERI used their CAD system to design complex dock templates. The bridge geometry was converted from Tekla to PERI CAD via IFC format; conflicts were discovered between anchor cable templates and tie rods at bridge anchor points. The design was modified to resolve these clashes (see Figure 8).

Figure 8: Collision between anchor cable and template pulling rod and its resolution

Construction Planning and 4D

The architectural model was first used in overall planning meetings and subsequently during reverse progress meetings as part of the Last Planner System (LPS). Vico Control™ software helped schedule on-site progress, with its visualization module used during planning meetings. After planning, schedules were imported into Tekla Structures v.15’s “Task Manager” window for detailed progress refinement.

The bridge deck was divided into two or three independent work zones to allow simultaneous operations by different teams. The model facilitated detailed planning of work sequences, quantities, and spatial information. Construction activities were assigned to model objects and highlighted with colors. Figure 9 shows a bridge deck section segmented into two work blocks, represented here in shades of gray.

Figure 9: Divided working sections (originally red and blue)

4D animation progress was updated daily, enabling the team to visualize and evaluate reverse progress during LPS meetings. This helped confirm whether planned work sequences were practical for space utilization. The animations improved understanding of tasks, enabling more precise work plans than traditional methods, and provided accurate spatial and quantity information.

Quantity extraction from the Tekla Structures model was straightforward, aiding material management by reducing excess stock and ensuring timely ordering.

However, due to manual linking of objects to activities, initial setup was time-consuming. When defects were discovered in central dock piles, construction was delayed by two months, requiring new piles to be installed. The team reversed the wall construction sequence to free time for rebuilding, working from both dock ends toward the center. The 4D model was not updated to reflect this change because redefining task relationships was costly and the schedule uncertain.

This experience highlighted the need for mature 4D software capable of defining high-level logical task relationships, allowing schedule updates with minimal effort and without redefining detailed tasks. At the time, Tekla Structures lacked this capability.

Manufacturing and Installation of Steel Components

The bridge model was shared with Ruukki, the steel manufacturer supplying components. Ruukki reviewed and modified components for manufacturing constraints, then sent updates to WSP and Skanska for approval. WSP integrated Ruukki’s feedback and updated the shared model.

Models were used bidirectionally to adjust design and production schedules. Ruukki synchronized with Skanska’s model to align manufacturing and delivery dates. Although internal data conversion between Ruukki’s model and resource planning software was manual, automation was anticipated. Construction progress updates improved material procurement, scheduling, and “pull” logistics for delivery and assembly.

Steel assembly onsite was handled by subcontractor Siltera, who occasionally referenced models for detailed production and process information, especially when drawings were unclear.

Reinforcement Refinement, Manufacturing, and Installation

Modeling steel reinforcement was more challenging than expected due to the bridge’s dense reinforcement and complex shapes. Unlike typical reinforced concrete structures with standardized elements, bridge components require customized modeling to accommodate curvature.

WSP was responsible for detailed reinforcement drawings, a contractual obligation for archiving and on-site use. Early collision checks avoided many spatial conflicts. Model data was also used to drive bending and cutting machines.

Tekla Structures produced quantity reports in ASCII, Excel, and other formats. For this project, ASCII report data was formatted for direct import into CELSA Steel Services’ manufacturing software, which controls NC machines in production (see Figure 10). This formatting effort, done in collaboration with CELSA, Skanska, and Tekla, eliminated manual steps and errors.

Figure 10: Steel bar data extracted from Tekla model and imported into manufacturer’s software

Skanska could not achieve the same integration level as CELSA, which had its own BIM-capable steel suppliers. ASCII files, specialized for steel bar shapes and quantities, were outside the scope of full model synchronization, requiring some manual data exchange. Therefore, steel bar tying and installation were planned externally and then integrated into the model.

The reinforcement workflow was as follows:

1. WSP uploaded design changes to Skanska’s model, refining steel bar details—a program bottleneck.
2. Skanska selected steel bars from the model based on construction progress.
3. Skanska generated reinforcement reports and sent them to CELSA.
4. CELSA imported the data into their software for manufacturing and delivery.
5. Skanska’s information personnel printed Tekla model snapshots for foremen to guide onsite assembly.

Steel bars were assembled onsite by Funnley, a specialist rebar binding company. Workers used paper drawings due to harsh weather and lack of modeling expertise. However, Tekla’s standard drawings were often insufficient due to the bridge’s complexity, causing friction among participants. Feedback was provided to Tekla for improved automatic drawing generation in future software versions.

Because drawings were sometimes incomplete, installers occasionally referred directly to the model for comprehensive views of steel bar arrangements. The contractor’s information personnel provided basic BIM training and assisted workers with model navigation and printouts.

Laser Scanning

Skanska hired an onsite inspector responsible for quality control and assisting contractors with positioning tasks. Initially borrowing equipment, Skanska later purchased a Trimble® VX™ Spatial Station, combining line-of-sight measurement and photography in one device, enabling solo inspections (see Figure 11).

Figure 11: Combined photos and scanned point clouds

Point clouds and images were uploaded to Trimble RealWorks software and compared with design coordinates converted from Tekla models, allowing real-time quality control of structures, templates, and embedded hardware. For instance, if an anchor cable’s actual position deviated by one meter, corrections were made before concrete casting.

The inspector participated in all planning meetings, including weekly Last Planner work meetings, helping identify technical requirements and constraints before work commencement.

BIM Supporting the Last Planner System (LPS)

Skanska Civil in Finland had three years’ experience with LPS before the Crusell Bridge project and had experts to train site personnel. LPS serves as a planning tool to identify feasible tasks by analyzing constraints, generating preparation checklists from weekly work plans. Key elements include reliable short-term planning and building a collaborative site community based on trust and commitments.

In this project, traditional LPS planning was followed with some modifications. Major suppliers and specialized contractors joined reverse progress meetings to develop 3-5 month work plans, creating an established task network. Models were used to visualize task composition and execution strategies. Site managers transferred results to Artemis PlaNet software for clarification and alignment.

Advanced scheduling followed, screening task constraints to minimize limitations, then establishing a three-week work schedule. This schedule was coordinated by Skanska’s construction team and subcontractors. The “Five Whys” technique was employed to identify root causes of tasks achievable only through LPS. Task completion was evaluated using Percent Plan Complete (PPC), averaging 84% with an 11% standard deviation.

Designers did not attend contractor-led LPS meetings, making steel bar refinement a bottleneck. The project manager suggested integrating designers into progress meetings to improve coordination. Construction personnel acknowledged ongoing learning needs with LPS, appreciating its benefits in understanding timing, flexibility, and obstacles.

When subcontractors missed planning meetings, production issues frequently arose onsite. Subcontractors sometimes provided unreliable task completion data and unrealistic delivery commitments—precisely the behaviors LPS aims to prevent.

6. Summary, Conclusion, and Key Learnings

Modeling reality in a virtual environment provided significant benefits to all Crusell Bridge project participants. In-depth BIM use improved management and organization, saving time and costs.

This case demonstrates BIM’s applicability to bridge projects. The team’s open and willing attitude toward BIM and lean construction methods (LPS) enabled learning from successes and failures. Their experiences strengthened future project delivery processes and contributed to software improvements. Complex projects naturally face challenges when adopting new methods, but resolving these leads to positive change.

Antti Karjalainen from WSP Finland noted, “The project’s results have been both positive and negative, forming a foundation for BIM development in bridges and software enhancements.”

Key takeaways include:

• Plan and implement BIM and LPS early: set clear goals, provide initial training, and foster a learning and improvement environment.
• Integrate models with construction management techniques such as planning, control, information exchange, meetings, and quality assurance.
• Utilize model synchronization for rapid information exchange.
• Use 4D scheduling to evaluate if the task network established in reverse progress meetings is feasible.
• When temporary structures are significant, model them accurately with correct quantities to enhance understanding during construction phases.
• Incorporate laser-scanned point clouds into models for efficient position verification and quality control, reducing rework.
• Leverage visual models in LPS planning meetings to improve comprehension of production processes.
• Encourage offsite project participants, such as the construction team, to regularly attend LPS meetings to synchronize detailed design and manufacturing information.
• Ensure all participants upgrade software simultaneously to prevent access and compatibility issues between different versions.

xuebim

Follow the latest BIM developments in the architecture industry, explore innovative building technologies, and discover cutting-edge industry insights.

← Scan with WeChat