In today’s post, we take a look at the efforts in the Latvian public sector to move towards a wider BIM adoption. In tow, we will take a look at the first Latvian national BIM pilot project ordered by SJSC State Real Estate and managed by BUVID and ITED.
In 2018, construction output in Latvia grew by an astonishing 21.9 %, which also contributed significantly to the annual GDP growth (4.8%). Among all the industries, construction was a top performer in value-added growth, a tendency that can be attributed to the inflow of EU funds inflow and other public and private investments in dwellings, other buildings and civil engineering.
The Latvian National Development Plan 2014-2020, the Transport Development Guidelines 2014-2020 and the National Roads Programme 2014-2020 all list the investment in the development and modernization of Latvia’s transport infrastructure, in particular roads and railways, as a priority.
It’s in this context that the Latvian public sector takes steps towards BIM. The following case study explains the background of why one of the biggest real estate owners in the country decided to procure BIM for all future projects.
The goal of the project was to assess BIM workflow in comparison to a traditional 2D workflow for a medium-size refurbishment. The scope of works included 2D-3D conversion, BIM coordination, quantity take-off and estimate comparison, 4D BIM and other tasks. The project delivery involved 11 companies and more than 25 BIM professionals. In what follows you can read about the project details as explained by the project participants themselves.
Location – Aspazijas boulevard 7, Riga, Latvia (map)
Employer – SJSC State Real Estate (Valsts nekustamie īpašumi, VNĪ)
Total project budget – 10.4 mil EUR
Gross floor area – 7 000 sqm (5 storeys and basement).
New function – prosecutor’s office building
Phase 1: Nov 2018 – Jan 2019
Phase 2: Jan 2019 – ongoing
The building is situated at the border of the Old Town of Riga and is 130 years old.
It was decided that the Pilot will be executed in parallel with construction works after the design had already been done. The process allowed to compare results and amount of issues identified with BIM and on-site. Consequently, solid indications regarding the potential amount of savings were revealed, in case BIM had been introduced earlier.
Deliverables, work packages and activities
The technical specification (TS) defined the tasks that had to be done and they were divided into three deliverables – BIM development, BIM analysis and As-built BIM development.
The first two had the duration of three months, and they are covered in this report. The third is still in process.
The main work packages (WP) were performed in the following sequence:
BEP & CDE
The first Pilot WP was to develop a BIM Execution Plan (BEP). The TS required models to be developed at BIMForum LOD300. LOD definition is a common approach to simplify the definition of what needs to be done. Though, it usually causes problems and Pilot was no exception – often defined LOD was insufficient for other activities, e.g. comparison of quantities. Such requirements were detailed and documented in the BEP. In parallel with BEP preparation, CDE was deployed. CDE solution was provided by the Latvian startup company OROCON.
The second WP was to create BIM models from 2D documentation of architectural, structural & MEP disciplines. Modelling timeframe was very short – 3 weeks. Few agile speed-improving decisions were made: 3 separate companies were developing an architectural model and the same amount – structural model. Revit work sharing is common in one office, but when it comes to 3 companies collaborating – drawbacks arise. IT support and private server configuration were required. As a result, models were delivered successfully after one building was modelled simultaneously by 8 different companies.
During the modelling stage, more than 100 design errors were found, missing or conflicting information. Many of these issues are quite impossible to make when 3D models are developed. For example, the same objects visible in one drawing were missing in the other, vertical pipe symbol locations were different between levels and other.
To support one of the BIM uses, quantities comparison, separate models for existing, demolition and new construction phases were developed. This also proved to be valuable during the laser surveying stage and spatial comparison of models with point clouds. In parallel, the design errors of information gaps were aggregated in a single report and presented to the employer.
The third WP included the creation of the federated model and performing BIM coordination. Two methods were used: walkthroughs/sectioning/filters and rule-based automated clash detection. Each discipline model was checked with itself and then with all other disciplines using Navisworks Manage.
In total, 40 issues were identified during visual coordination and 180 during automated clash detection. Most of the issues were replicated across the building, and one group could include hundreds of similar clash instances across the building, for example, pipe clashes in the same model.
A simple estimate shows that dealing with these issues on-site costs a significant amount of money. The assumption was made that one issue can be resolved by 2 people in 2 days. In total it results in 7104 h spent. With an hourly rate of 20 EUR solving these issues would cost more than 140k EUR.
The fourth WP included the comparison of the bill of quantities developed during the design stage with quantities stored in BIM. Several assumptions were made:
1. Models precisely represent the design solutions
2. It is impossible to compare all of BOQ positions with BIM. BIM and design were with different detail levels.
3. Only physical volumes were compared. Labour costs and effort were not tested.
Autodesk Revit was used – generated schedules in separate models, that correspond to BOQ structure, were exported to MS Excel.
Despite the mentioned assumptions, several important conclusions were made:
– Human errors can be reduced significantly by using structured information
– Data Science can be used to make better decisions which are based on information stored in BIM
– The average delta for quantities between models and BOQ exceeded 10%
The fifth WP was to develop 4D simulation by adding confirmed construction schedule data to BIM elements. The process showed that provided schedules are uninformative, under detailed and virtually unusable for construction works planning. The video below shows how interior works would be done. The exterior walls are made transparent.
This WP was initially out of scope, and it was delivered for free to showcase the potential of laser surveying during construction. It was decided to scan after all the demolition works are done. Developed architectural and structural models also included this phase, and it was feasible to compare the differences between reality and BIM developed based on the original design.
The Pilot was intended to show what are the benefits of BIM for the Latvian construction industry, especially for state and public clients. It can be admitted that this goal was achieved. Here are the main conclusions from the Associaton for Construction Industry Digitalisation of Latvia:
– It is expensive not to use BIM. 10-20% costs could be saved by better design, coordination and planning.
– Traditional 2D methods are insufficient. It is nearly impossible to coordinate design without 3D models.
– The Pilot defined only a few BIM uses. In reality, there are hundreds, that can deliver value to the projects.
– Laser survey is an obligatory technology for reconstruction projects. The initial costs are outweighed by the avoidance of losses due to design errors.
– Great quality and results require strong competence on both sides – the client and the supply chain.
The Pilot proceeds with as-built model development. More conclusions will be made and we will write about them.