Sustainability in Structural Design in High Seismic Regions

To view the figures and tables associated with this article, please refer to the flipbook above.

With the growing focus on climate change and the pressing need to reduce construction-related carbon, the need for sustainable structural design is imperative. According to a 2019 World Green Building Council report, the built environment contributes 39% of global carbon emissions, with 11% from embodied carbon, which includes emissions from manufacturing, transport, construction, and end of life phases of the built environment. According to the studies, this percentage could rise to 50% of new construction's carbon footprint by 2050. Thus, a shift from building new structures to retrofitting and retaining existing ones is essential to minimize embodied carbon. Minimizing upfront and embodied carbon is the primary “focus."

In seismic regions, minimizing embodied carbon while ensuring seismic resilience presents a major challenge. Conventional ductile-design strategies often lead to significant damage and rebuilding after earthquakes, as seen in events like the Christchurch earthquake, where about 70% of the buildings in the Central Business District had to be demolished (costing NZD $45 billion, which amounts to approximately 20% of New Zealand’s GDP). To retain existing buildings in such high seismic areas, a completely different approach is required. Retrofitting existing structures with solutions like viscous dampers is yielding cost-effective retrofit solutions and significantly enhanced seismic resilience.

Viscous dampers are highly effective in reducing earthquake impacts on buildings in seismic regions. These fluid-mechanical devices generate forces proportional to velocity, counteracting structural displacements and accelerations. If properly designed, they reduce both drift and floor acceleration, helping to minimize building damage and downtime, providing enhanced resilience during and after an earthquake. Especially for retrofit, they are more efficient than other technologies like base isolation, as they require less extensive foundation work and superstructure stiffening which is often very costly. The key to their (viscous damper incorporated retrofit) effectiveness, however, lies in the quality of the retrofit design. The key question that arises is what do we really mean by “if properly designed” and why is it important to include this qualification?

A typical section of a viscous damper is shown in Figure 1.

Equation 1 in the sidebar "Mathematical Modeling of Viscous Dampers," describes the dynamics of a building equipped with viscous dampers during an earthquake. This equation may be visualized as a triangle, with three pivotal figures occupying the corners, symbolizing the key forces involved: Sir Isaac Newton (representing inertia), Robert Hooke FRS (representing the nonlinear restoring force), and Lord Rayleigh (representing damping). Together, they form the left-hand side of the equation (1). When an earthquake strikes a building, an inherent “debate” takes place on how the total load will be distributed among them—essentially, how much resistance each will contribute to counter the imposed earthquake forces. This complex interaction, or "load sharing," makes the problem of earthquake engineering highly complex. Adding mechanical devices like viscous dampers only amplifies this complexity, as they introduce additional layers of resistance and force distribution.

Traditional methods of design for conventional structures (structures without dampers) incorporate numerous simplifications to this phenomenon of “load sharing” mainly for mathematical convenience.

Typically, the “proper design” of a viscous damper incorporated structure adopts the Performance Based Design (ASCE41/FEMA P58) using nonlinear time history analyses. This involves subjecting the proposed design to a suite of ground motions (usually 11 bi-directional ground motions) and establishing that it satisfies the relevant compliance requirements and target performance criteria. Variations in material properties also need to be factored into the design.

The standard industry approach is to extrapolate the simplified design approaches normally used for conventional structures to those incorporating viscous dampers. This results in an initial starting point. From there, we have observed engineers reverting to an intensive "brute force" process, where multiple analyses (in the range of 1,000s) are conducted to refine the system. This involves varying damper locations and parameters in an effort to optimize the design. Depending on the building size and complexity, this iterative process can take months or even years to arrive at a compliant design.

Each iteration requires significant computational effort. For one intensity level, a single iteration involves 22 nonlinear time history analyses. Since there are usually at least three intensity levels (Serviceability, DBE, and MCE), each with specific performance requirements, this means that one iteration realistically entails 66 analyses, if all intensity levels are considered simultaneously. Depending on the building's complexities, multiple intensity levels may need to be considered simultaneously during the design process and hundreds to thousands of iterations may be necessary, making the process computationally intensive and extremely time-consuming.
As a consequence, engineers have needed at times to terminate their efforts once they have achieved what they deem to be a compliant design as opposed to persisting in pursuit of an optimized design that consumes less material and achieves superior performance. The resulting designs tend to be material intensive, driving up cost and carbon emissions, while often falling short of achieving optimal performance.

Now why does this happen? The primary reason for this inefficiency lies in the "brute force" approach itself. In this method, engineers lack the ability to assess the complex load-sharing process that is crucial for determining the ideal placement and sizing of the dampers. Without a clear understanding of how loads are distributed across the system, it becomes nearly impossible to strategically quantify and position the dampers. This lack of insight leads to a more labor-intensive design process that ultimately falls short of optimizing performance.
To address these challenges and facilitate smooth commercially viable applications of Performance Based Design approaches (ASCE41/FEMA P58), a proprietary smart design platform called MOODD, based on advanced PBSD principles was developed. This platform produces design solutions explicitly accounting for these intricate physical interactions. By optimally tuning these “load sharing” interactions, the platform generates designs that not only enhance seismic resilience for a given cost, but also promote sustainability through reduced upfront carbon emissions. In addition to PBSD, the platform also has the capability to generate designs based on Performance Based Wind Design (PBWD) principles where the wind dynamic loading is considered which maybe applied for tall and super tall buildings. Two real life applications of the platform for retrofit for seismic loading are described in this article.

8 Willis Street is in the Central Business District of Wellington. The city rests on the meeting point of two tectonic plates and is prone to large earthquakes. Major fault lines very close to the city include the Wellington Fault, the Wairarapa Fault, and the Ohariu Fault. The Wellington fault may be able to generate earthquakes larger than magnitude 8.0 and may generate ground velocities around or larger than 1.5 meters/second and ground displacements greater than 2 meters. Obviously a retrofit design in Wellington should cater to these sorts of extreme seismic demands.

The seismic rating of the existing structure built in the 1980s (Fig. 2) had fallen below what is acceptable for a commercial office space. To improve the commercial viability and future-proof of the existing asset, property developer Argosy, headquartered in Auckland, New Zealand, decided to retrofit the building to be 130% New Building Standard (NBS) along with increasing the floor area from 6,500 square meters to 11,750 square meters.

The primary structural system of the original building was comprised of reinforced concrete moment frames in one direction and shear walls in the other direction. The floor system was precast hollow-core concrete floors, and the foundation system was shallow pads.

Recent earthquakes like Kaikoura (2016) revealed the vulnerability of hollow-core flooring systems during earthquakes. The main issues relate to loss of seating (where a unit slips away from its support), positive moment failure, negative moment failure, and web splitting which are all drift/rotations related very similar to issues exhibited by pre-Northridge connections. These issues were all present at 8 Willis Street.

The increase in floor area was achieved by adding five stories on the top of the existing eight stories and a 13 meter extension to the street front (Fig. 4). Figure 5 shows the retrofitted and enhanced 8 Willis Street building in the present day.

Performance based seismic design using the smart design platform was applied for the retrofit design incorporating fluid viscous dampers. The primary retrofit scheme includes twelve viscous dampers arranged in the moment frame direction. The fluid viscous dampers are tuned in such a way that the overall building performance met the target requirement of the client with no additional foundation work for seismic loading. Seventy ground motions scaled to NZ 1170.5:2004 were used for the design. Aleatoric and epistemic uncertainties were explicitly accounted for in the design. An independent peer review of the entire retrofitted structure was done by the world-renowned Earthquake Structural Engineer and Structural Dynamist Prof. Emeritus Athol J. Carr (author of 3D Ruaumoko software) from University of Canterbury, New Zealand.

The final retrofit had more floors without dampers than with dampers and met the Architectural requirements of having uninhibited floor plates for maximum functionality. A typical arrangement of fluid viscous dampers is shown in Figure 6.

The adaptive reuse of the structure alone resulted in saving a whopping 1904 tons of carbon.

33 Bowen is a 12-story reinforced concrete moment-frame building designed and constructed in the 1980s, incorporating precast hollow-core floors (Fig. 7). The moment frames were well-designed and detailed for expected design level according to NZ 4203:1984 which is the predecessor of the current New Zealand code NZ 1170.5:2004. Figure 7 shows the existing side elevation of the building.

Beca's seismic assessment rated the building at 35% NBS, just above "earthquake prone" status, with the low rating due to hollow-core diaphragm issues and the frames nearing code drift limits. Coupled with shear failure concerns from beam overstrength not accounting for slab reinforcement, the deficiencies identified led to tenants vacating. The situation was further exacerbated by new GNS Science data indicating a 90% hazard increase for the site.
The commercial chaos caused by the tenants vacating due to the low seismic rating made the owners decide that the retrofit needed to account for potential future changes to the loading standard that are anticipated to incorporate the updated National Seismic Hazard Model. It was decided that for future proofing, 80% of new seismic hazard will be the target. This sets the target strengthening for the existing building at 150% NBS based on current hazard.

To achieve the target of 150% NBS, nearly a four-fold increase in seismic resistance was required. Even to achieve 70% NBS (twice the current rating of 35%) through conventional means of strengthening would demand an enormous amount of intervention, both in the foundations and in the superstructure. So, achieving 150% NBS through traditional means of strengthening was impractical from a cost perspective with an associated negative consequence from a carbon perspective. In addition to financial infeasibility, this intervention will also affect the functional and aesthetic aspects of the building, rendering it un-lettable. To ensure an attractive commercial outcome, a smart retrofit strategy was necessary—one that would deliver significantly higher performance with minimal intervention and at a more acceptable cost.

Viscous dampers were chosen for the project due to their "out of phase" response to structural forces. However, to achieve the desired performance, the retrofitted system had to be “properly designed” considering the “load sharing” phenomena. Traditional design methods, which conveniently and incorrectly ignored these effects, couldn't meet the project's objectives of performance and cost. The client only considered the project to be financially/commercially feasible, if 150% NBS rating could be achieved at a cost which is less than the cost of conventional strengthening for a target of 70% NBS. To achieve this, the PBSD using the smart design platform was employed to generate the design. The obtained design was then evaluated using the NZ 1170.5:2004 alternative solution compliance pathway to demonstrate compliance.

Figure 8 shows the ADRS (Acceleration-Displacement Response spectrum) curve for the existing structural system, the drift of the uncontrolled (without dampers/original structure) was around 9% in MCE earthquake (blue line in Figure 8) and 5% in the DBE (green line in Figure 8). The target drift was chosen as less than 1% in DBE (a five-fold reduction) and less than 2.0% in MCE (a four-fold reduction) to mitigate the hollow-core issues as per the recent experimental research guidance from SESOC (The Structural Engineering Society of New Zealand). To put these drift targets in perspective, it is worth noting that the standard compliance drift target for NZ codes is 2.5% for DBE.

Figure 9 represents the drift plots for both longitudinal and transverse directions of the building when subjected to multi-directional earthquakes. Maximum inter-story drift in DBE is around 0.8% and the maximum inter-story drift for MCE is <1.8%.

Implementing the smart design platform at 33 Bowen dramatically reduced on-floor work, making the retrofit process far more cost-effective than conventional design approaches. As shown in Figure 10, the comparison between traditional and PBSD designs using this advanced framework reveals that the design solutions incorporating the “load sharing” effects of physics results in significantly more sustainable solutions with minimal initial carbon footprint. Moreover, the smartly designed viscous dampers negated the necessity for any foundation work due to substantial reduction in the base shear.

Figure 11 illustrates the viscous damper arrangements in 33 Bowen. The damper locations are confined to the perimeter frames to avoid introducing any obstructions on the open-plan floor plates.

The article highlights the effectiveness of combining viscous dampers with advanced structural design technologies, such as the smart design platform, to identify optimized retrofit solutions. This approach has proven successful in enabling cost-efficient seismic retrofitting of buildings with significant vulnerabilities (drift and acceleration related) in high seismic regions. The resultant retrofit solutions achieved a substantial reduction in total cost and embodied carbon compared to that possible using conventional analysis and design techniques and obviously even greater savings when compared to the alternative of the demolition and reconstruction of a replacement building. ■