TO SUGGEST THAT THE DESIGN OF THE CALIFORNIA ACADEMY OF SCIENCES (CAS) IN SAN FRANCISCO EVOLVED FROM AN INITIAL CONCERN ABOUT EVIDENCE-BASED DESIGN WOULD BE ERRONEOUS AND MISLEADING. NEVERTHELESS, THE CAS IS A PERFECT EXAMPLE OF GREAT OUTCOMES THAT CAN RESULT WHEN DESIGN CREATIVITY COMES TOGETHER WITH RESEARCH TO ENHANCE BUILDING PERFORMANCE.
The CAS design process utilized surveys and observations like those often used in social scientific research, as well as computer modeling, simulation, and full-scale mockups. It drew on design architect Renzo Piano's experience, as well as project-based research, and it took maximum advantage of interdisciplinary expertise from subconsultants and subcontractors. These methodologies all characterize an informed design process.
The CAS client and design team didn't state they wanted to follow an evidenced-based process, but they chose to fully understand the implications of many complex design choices that needed to be made. To this end, they utilized many of the methodologies described in previous chapters of this book, especially survey, simulation, and modeling. Utility, sustainability, constructability and delight resulted. (See Insert C2)
The design by Renzo Piano Building Workshop (RPBW), in collaboration with Chong Partners Architecture (now Stantec), ARUP, and Webcor Construction, began with an inspired design vision created by Renzo Piano. This vision grew from the site and a spatial interpretation of the Academy's multi-faceted mission. Form and function. The design parti formalizes cultural, historic, and environmental context through the lens of Piano's personal design genius and experience. (See Insert C1)
Piano is equally concerned with "the making" of a building and the strength of its initial design concept. The five-phase design process typically used in the United States progresses from schematic design through design development, construction documentation, bidding, and construction administration. Most creative activity occurs in schematic design, with subsequent phases being predominantly about technical implementation. Design is generally allocated only 35 percent of the total effort.
In contrast, Piano's process is iterative. He starts with a vision and then almost immediately turns his attention to structure, environmental aspects of building performance, construction technology, and materiality. Each aspect of design contributes to the design concept and the enhancement of his initial vision. DESIGN IS EVIDENT IN ALL PHASES OF WORK, AS ARE ATTENTION TO DETAIL AND USE OF EVIDENTIAL PROCESSES. It is this process of working from "whole to part to whole," moving from creative intuition to evidence and then back to the vision, which results in extraordinary design innovation. Seeking and applying evidence enhances creativity, as concept evolves to reality.
Piano's beginning with a vision of what might be, based on personal intuition and experience, is quite similar to the way a scientist creates a hypothesis. The past suggests what might come next, but the idea springs from creative thought. Just as Piano then uses experimentation to develop the elements of the design, a scientist pursues his/her research plan using a variety of empirical methods. Designer and scientist both look to some form of valid and reliable evidence, whether quantitative or qualitative, for assurance that the work will result in successful outcomes.
Rusty Gage, neuroscientist from the Salk Institute, first drew the author's attention to the similarities of how architects and scientists use evidence. Gage finds the similarities to be apparent. But architects continue to fear that using scientific methods will diminish their creativity. Thus, they dismiss the use of an evidence-based design process.
THE CAS STANDS AS AN EXEMPLAR THAT GREAT DESIGN RESULTS FROM CREATIVE VISION AND CREATIVE USE OF EVIDENCE. Much of the evidence this case study will cover comes from computer simulation, modeling, and full-scale mockups (a form of modeling and rapid prototyping as a principle methodology of generating evidence). In addition, at the project outset, the Academy staff utilized social scientific methodologies to develop evidence to inform functional and operational programming. These included user surveys, benchmarking, and field investigations of the existing facility and other scientific museums.
Finally, because of the manner in which these design explorations were undertaken, CAS is fortunate to have a wide range of sensors that provide a plethora of data that is continually being captured by a computer facility management system. This "feedback loop," described further in the case study, provides another form of evidence, to be used to improve building performance on an ongoing basis and to inform future projects by the CAS team and others in the design and construction professions.
The California Academy of Science originated in 1853 and relocated to Golden Gate Park in 1914. The CAS has prospered from being a unique combination of natural science museum, aquarium, and planetarium under one roof. Its mission is even broader. A research and educational institution, the CAS's stated mission from its inception has been "to explore, explain, and protect the natural world." The building design functionally and aesthetically supports this mission; the process of design was exploratory and scientific.
From 1914 to 1989, CAS expanded into 12 separate buildings, located on the same site as the new building—the Music Concourse of Golden Gate Park. Damage from the 1989 Loma Prieta earthquake forcedclosure of portions of the museum, driving a decision by the board of directors to build a new Academy. In 1999–2000, the board hired Renzo Piano and began the design process. The board expected intense public scrutiny of the design and knew that every decision would have to be backed up.
By contractual agreement, Piano and THE CAS LEADERSHIP TEAM COMMITTED TO A PROCESS OF EXPLORATION AND INNOVATION IN ALL AREAS OF DESIGN, ESPECIALLY IN THE REALM OF SUSTAINABILITY. For the CAS, a scientific institution, a process of inquiry was the logical approach.
The CAS reinterpreted what the "standard of care" for their architect would be, when requiring an exploratory process. They also backed up their intentions by agreeing to pay for nontraditional investigations, including full-scale mockups well beyond the what would be within the traditional scope of design and construction services. These evidence-seeking processes are integral to Piano's commitment to innovation and resulted in benefits for the CAS and the surrounding community. For their part, the leadership of the California Academy of Sciences was insightful, bold, and creative in support of scientific and design innovation using evidence.
Some architects claim that each building that is designed and constructed is a prototype for the next. Unlike an auto manufacturer, as one example of a product developer, which undertakes a lengthy period of research, development, and prototyping prior to letting the first car roll off an assembly line, an architect generally first sees a complete physical manifestation of his/her design creation when construction has been completed. By then, a considerable investment has been made without knowing if the outcome will be of value. Given their fiduciary role, the CAS board's agreement to permit Piano's use of full-scale mockups was clearly a wise investment. (See Insert C3–C5)
The project is approximately 410,000 square feet. It includes:
Exhibit halls
An auditorium
Retail shops
Food service
Three iconic exhibits for a planetarium, a self-contained rain forest, and an aquarium
Approximately two acres of landscaped roof used for educational purposes
Two weather stations located on the roof
Photovoltaic panels generating 5 percent of the building total electrical use
approximately 200,000 square feet of research laboratories, offices, and collection storage
The design and construction took approximately eight years and cost approximately $480 million. Given this complexity and investment, it's apparent why the client needed to have confidence that the right design decisions were being made at every step of the process. Many types of evidence were sought and applied, all benefiting the project's success.
Immediately prior to hiring Piano, then-Executive Director Patrick Kiocelek, PhD, led an 18- to 24-month effort of defining what the CAS should be. As a part of that process the Academy conducted over 150 focus group meetings with various stakeholders and experts to help define the future of the Academy for the twenty-first century. These efforts utilized traditional social scientific surveys, questionnaires, in-field observations, and benchmarking of other comparable scientific facilities. Diverse, interdisciplinary stakeholders participated in this effort to ensure that the institution's mission would be addressed from multiple perspectives. This highly systematic and thorough process assured that the correct questions were asked of a representative sampling of people and that, thereby, a compelling design hypothesis would form.
Armed with this knowledge, specialized museum consultants were hired to substantiate the reasonableness and feasibility of the strategic operating model, or hypothesis. These consultants provided the equivalent of a peer review, a step within scientific research protocols. From the survey research and peer review process, a substantial body of evidence was gathered and became the basis for the functional, experiential, operational, cultural, educational, scientific, and philosophical strategy for the new building.
This evidence framed the design program in quantitative and qualitative terms. With this information in hand, combined with cost parameters, clear process mapping, and definition of expectations, the design team had a solid basis to understand what they needed to do.
Evidence produced through simulation modeling and full-scale mockups had three intended purposes:
To support, refine, and build confidence in the design; specifically the sustainability strategies and structural approach. Permit the team to enhance the design vision by establishing a better understanding of building performance and expected outcomes.
To support technical code compliance, construction detailing, sequencing, and budget refinement.
To have an ability to monitor actual system performance and to adjust the systems to enhance operations.
ARUP was the integrated engineering firm responsible for the design of structural, mechanical, electrical, and plumbing; facade consulting; acoustics consulting; sustainability consulting; lighting design; and life-cycle analysis. Their London office worked closely with Renzo Piano Building Workshops on the initial concepts and collaborated with ARUP's San Francisco office in the simulation modeling and mockup development.
The use of performance-based code compliance, permitted under the prescriptive code's "alternative materials and method of design" clause, was uniquely allowed by the San Francisco Building Department and was used on more than a dozen such performance-based approaches to seismic design, fire safety design, and energy conservation. Although effective for analyzing complex structures, performance-based compliance can be labor intensive and require specialized expertise, as well as more fee and time for design and agency reviews. Prior to beginning the research, it's necessary for the design team and reviewing agency to agree on the assumptions, design methodology, and evaluation criteria. Reviewing agencies will often employ the expertise of a third-party peer review consultant.
For the CAS, ARUP provided innovative research efforts, focused on Computational Fluid Dynamics (CFD) modeling of the natural ventilation system; occupant flow analysis with simulation modeling; and development of a lighting system for the live penguin exhibit, which mimics the lighting conditions of their home environments in South America.
One example of the research is the calculation of (natural) ventilation effectiveness of the exhibition hall. This study confirmed the feasibility of naturally ventilating the exhibit hall and thus obtained LEED points derived from Credit EQ-2.
The exhibition hall is a large, crossed-shaped space, approximately 424 feet long by 248 feet wide and 61 feet tall at its maximum height. The hall has two large domes, the planetarium and the rainforest. Both are mechanically ventilated and self-contained. An external piazza is situated at the center of the exhibition hall. All three spaces are represented as obstructions in the model.
The ability to naturally ventilate this large space eliminated a significant consumption of energy, which would have been necessary to mechanically ventilate the space. THE IMPACT OF THIS STUDY IS SIGNIFICANT TO THE ONGOING FISCAL IMPACT OF THE OPERATION AS WELLAS TO THE DESIGN CONCEPT.
A CFD analysis was carried out as a means of calculating the mean age of air and the air change effectiveness at all points within the modeled zone, as described in ASHRAE Standard 62-2001. No-wind conditions represent the worst-case scenario, when ventilation relies principally on buoyancy and/or stack effect. The CFD model provides a numericalanalysis of the worst-case conditions for both cold winter mornings and hot summer afternoon conditions.
The natural ventilation in the exhibition hall depends on open windows at high and low levels on each entrance facing the hall, to provide fresh air into the space, as well as on high-level skylights above the domes of the rainforest and planetarium, for exhaust of warm air. On a design day summer afternoon, all windows on each facade are fully open, as are operable skylights on the roof, corresponding to a free open area equivalent to 13.5 percent of the overall facade. The radiant floor is then set in a cooling mode.
On a design day winter morning, only the windows at high level are partly open (10 inches) and some of the skylights above the rainforest are open, corresponding to a free open area equivalent to 2.2 percent of the overall facade. The radiant floor is then set in a heating mode. (See Insert C6–C9)
A literature search confirmed that CFD is often used in the design of mechanically ventilated and naturally ventilated spaces. It was also determined that the degree of ventilation effectiveness is defined by a function of age of air, nominal time constant, and the arithmetic average of the ages of air measured at breathing level within the space. From this literature search, confirmation of the methodology (CFD) and the metrics were ascertained from the beginning.
Nine specific steps were taken to perform a virtual tracer gas decay test using CFD. The nine-step procedure was applied for summer and winter conditions and calculations made of age of air, nominal time constant, and air-change effectiveness.
The methodology included:
A literature search confirming the use of CFD and the choice of metrics.
Specific procedures and controls
Stated assumptions and definition of worst-case conditions
Testing over a prolonged period resulting in case studies for summer and winter
The results are presented in degrees Farenheit and compare summer and winter conditions, in terms of average, maximum, and minimum temperatures, as well as average and maximum air speeds. The ventilation effectiveness is assessed for both summer and winter against concentration and air flow at exhaust air streams, age of air at breathing levels, nominal time constants, and air-change effectiveness.
The CFD analysis indicated that for the exhibitionhall, the proposed ventilation strategy would maintain temperatures within the design comfort range for peak summer and winter conditions, provide ventilation higher than 0.9 and meet LEED Credit EQ-2. Additionally, the CFD analysis provided a means to prove a performance-based code compliance for the space to be naturally ventilated.
To satisfy Title 24 criteria, the same analysis technique was applied to a typical office space where ventilation consists of a mixed system with natural ventilation in the areas closest to windows and mechanical cooling and heating for the areas more than 24 feet from the perimeter.
The results of the study proved the effectiveness of a naturally ventilating strategy for the exhibition areas, developed ongoing operating expectations of the space, provided performance-based code compliance, and informed designers as to size, location, and number of openings related to air intake on the facade and exhaust from the skylight design.
For operational reasons, the collection storage area needed to be in close proximity to working scientists, as opposed to a remote location. The collections include approximately 6 million scientific specimens, the majority of which are in "wet containers" filled with alcohol and ethanol. The total amount of flammable liquid (estimated at 350,000 liters in thousands of these glass containers) formed a serious concern for a public assembly building. Complicating the issue, the specimens are stored in a high-density mobile storage (compactor) system.
In addition to a need to create multiple 4-hour separations, the challenge was to research and test the feasibility of an automatic fire suppression system that would satisfy code requirements for secondary containment of spills, provide fire suppression water, and not damage the specimens.
The investigation methodology included a survey of alternative, nonconforming building solutions. This search led to examination of a high misting system more commonly used in aircraft carriers and oil tankers for fire suppression. The use of this concept was tested, peer-reviewed by an expert, and validated with full-scale mockups built in Texas. Three fire tests were performed which confirmed the performance of the misting system and, because of their low flow rates, eliminated the need for a special drainage and containment system.
A similar process was used to address the dry specimen collection. In this case, an alternate solution requiring a custom-designed sprinkler system was developed and tested in Canada.
WITHOUT THE PERFORMANCE APPROACH, THE USE OF FULL-SCALE MOCKUPS, TESTING, PEER REVIEW, AND USE OF PUBLISHED TESTING REFERENCES, THE DESIGN AND FUNCTIONALITY WOULD HAVE BEEN COMPROMISED, THE AMOUNT OF STORAGE WOULD HAVE BEEN LESSENED, AND RESEARCH ACTIVITY WOULD HAVE BEEN IMPAIRED.
In addition to these two specific examples of use of evidence for performance-based design, ARUP effectively informed and enhanced the design with studies permitting unprotected exterior steel beams, a three-story ramping for the rainforest's spatial expression, and a uniquely designed acoustical ceiling solution. Each of these was in conflict with the prescriptive code but shown to be code compliant as a result of intensive research and creative investigative approaches.
THE ENGINEERING SOLUTIONS, WITH EXAMPLES NOTED ABOVE, EMPHASIZED THE USE OF LITERATURE SEARCHES, SCIENTIFIC TESTING METHODOLOGIES, COMPUTER SIMULATION AND MODELING, AND PEER REVIEW AS APPROACHES TO EVIDENCE-BASED DESIGN AND PERFORMANCE-BASED CODE COMPLIANCE. SIMILAR BUT DIFFERENT APPROACHES TO SEEKING EVIDENCE WERE USED FOR THE ARCHITECTURAL DESIGN ELEMENTS. THESE WERE DONE, AS WERE THE ENGINEERING STUDIES, WITH THE INTENT OF INFORMING DESIGN, ENSURING PERFORMANCE-BASED CODE COMPLIANCE, AND VALIDATING CONSTRUCTION FEASIBILITY AND DESIGN DETAILING.
The architectural process included the following:
Translating design sketches from RPBW into 2D and 3D computer drawings to reveal interfaces and aesthetic impact
Development of architecturally scaled models to view scale and proportion in 3D
Full-scale, offsite mockups by manufacturers of the subsystems including:
Roof structure framing
Roof canopy with photovoltaic cells
Various window wall and mechanized shading devices
Landscape roof system: soil erosion, slope stabilization, irrigation, selection of plant material, and installation processes
Steel tensegrity and glass system covering the piazza
In some instances, there was an additional step of using full-scale mockups made of wood, prior to the use of the selected material. These helped develop refined architectural decisions related to scale, proportion, light transparency, etc.
In yet other instances, onsite full-scale mockups made of the specific material followed the offsite mockups created by the foreign manufacturers. These were tested for integration with other structural, architectural, or construction systems. If a specific system, such as the Glass Dome (Bolla), was a third-generation design previously used by Piano, fewer steps and mockups were necessary for design reasons. Prior experience, in these cases, provided adequate evidence for design. However, even in these cases, mockups were used for code compliance, testing agency, and subcontractor understanding.
THE USE OF FULL-SCALE MOCKUPS AS A WAY TO CREATE EVIDENCE BRINGS TOGETHER AN INTERDISCIPLINARY PERSPECTIVE TO INNOVATION. It engages the science of construction technology with the physical sciences, in performance-based code compliance. The CAS design process drew evidence from many sources, testing the most complex of situations using various approaches and addressing a variety of hypotheses. (See Insert C10–C12)
Beyond informing design, the use of full-scale mockups reduced risk by increasing understanding and reducing first cost. The general contractor, Webcor, and its specialty subcontractors reviewed the mockups and provided critiques. This savings is evidenced by the fact that the successful low bidders were those who participated in the creation of the mockups. The bid spread between those who understood the evidence and those who didn't was an indicator of actual construction cost savings.
Most buildings are designed such that the interior environment is independent, and not reflective, of what is happening on the exterior relative to heating, cooling, air movement, and light. The building "skin" hermetically seals and isolates the interior environment from the impact and changes to exterior conditions. Thermostats and manually operated light switches determine thermal comfort for the occupants and dictate the required mechanical and electrical system functions in response to heating, cooling, and electrical needs.
Unlike most buildings, the CAS is designed such that there is a strong relationship between the exterior thermal environment and the resulting interior environment. Similarly, interior occupant and thermal conditions can trigger changes to the building's facade, which is sometimes open to light and air and sometimes closed. THE RESULT IS A BUILDING WHICH IS DYNAMIC, RATHER THAN STATIC—ITS BUILDING SYSTEMS (ELECTRICAL, HEATING, COOLING), SKIN, AND SKYLIGHTS CONSTANTLY ADJUST IN RESPONSE TO CHANGING EXTERIOR CLIMATIC CONDITIONS.
Starting with information received from the two weather stations on the roof, the computer facility management system receives the information about the exterior and integrates it with information about the interior environment received from sensors. Signals are sent to actuators that cause the building openings, including skylights, windows, and shades, to open or close in response.
This highly automated system results in a well-documented understanding of building performance that is supported by the USGBC LEED process. This process results in three unique and distinctive operational conditions:
Balance: A constant balancing of building systems to coordinate between dynamic interior and exterior conditions is needed.
Integration and Interface: Green buildings, which depend upon more automation, require greater design integration of systems and controls, as well as attention to the interactions among subsystems. The current construction model that defines the roles and responsibilities of the general contractor and his/her subcontractors doesn't adequately address the interface and integration in highly automated buildings. The design and installation supply chain needs greater attention to work smoothly.
Data: This system-balancing process requires use of sophisticated technology and management skill but results in an operation that is "data rich."
HAVING A HIGH VOLUME OF HIGH-QUALITY DATA HAS PERMITTED CAS MANAGEMENT TO MAKE SYSTEMS-RELATED OPERATIONAL AND FISCAL DECISIONS THAT MAXIMIZE THE TOTAL BUILDING PERFORMANCE. Of particular value, the data has been provided in "real time" thus permitting an immediate reaction from the feedback loop. In addition to being able to isolate performance of individual systems, the data can be seen in comparison to other functions that inform the interfaces, interlock, or action and reaction of dependent systems. It is anticipated that once an adequate amount of baseline data has been developed, the data will be used to forecast and anticipate facility needs rather than merely react.
The use of building performance data collected at the CAS is considered to be part of a phased strategy for continual improvement. The CAS facilities management team, led by Ari Harding, has developed a high level of sophistication to monitor, refine, and improve the continuous operation of the facility systems. Getting good feedback on the system operations is an important part of the use of evidence and data in design and operations. Harding identified the following stages of work for their program:
Phase 1A: Approximately a year following completion of construction, most building systems were being run and tested during installation of exhibits. However, complete system functions could not be tested until the building was fully occupied with staff, equipment, exhibits, and visitors in place.
During the nine months since the September 2008 opening, the commissioning efforts have been focused on adjusting and calibrating the systems to work as designed.
Phase 1B: During this same nine-month period, corrections, additions, and resolution of issues raised during commissioning have taken place. CAS management acknowledge that for a building as complex as this, perfection of all systems and their interfaces is not the appropriate expectation. Although corrections, additions, and resolution of issues made during ongoing operations is challenging, the need to adjust over time is reasonable.
Establishing a Baseline
It is anticipated that formal commissioning will be completed by the end of one year of occupancy (September 2009). From this time forward, major systems should be operating as desired and baseline data will provide reliable, consistent, and dependable data.
Comparative Analysis
A comparison between actual system performance against the optimum "Gold Standard" will set clear performance metrics.
By September 2010, a full year of operations will have provided seasonal insight and context for the benchmarked data. After this, it's anticipated that a comparative graph of actual performance versus optimum performance will be charted for all systems.
At this point the data is stable and benchmarks are known. CAS managementcan move from a reactive mode of using data to make corrections when needed, to using thedata as a forecasting and modeling resource based upon trend lines.
IN ADDITION TO USING DATA AS A MEASURE OF UNDERSTANDING BUILDING PERFORMANCE, CAS MANAGEMENT WILL BEGIN TO USE SURVEYS, QUESTIONNAIRES, AND FIELD OBSERVATIONS TO DETERMINE (STAFF) OCCUPANT SATISFACTION AS ANOTHER FORM OF EVIDENCE. Again, this survey process is in keeping with green building management principles. (See Insert C13–C15)
Isabelle Lavedrine (OVE ARUP & Partners, California Ltd.): Computational Fluid Dynamics (CFD) Modeling of Exhibit Hall and Typical Research, Collection and Administration (RCA) for LEED Credit EQ-2 : Design Study, 2007.
Armin Wolski, Geza Szakats Susan Lamont (OVE ARUP & Partners California Ltd.): Margaret Law Award 2008 San Francisco Submission—Fire Engineering, Design Study 2007.
Kang Kiang, Project Manager on the California Academy of Science for Chong Partners. Personal interview with Kang Kiang, june 2009.
Ari Harding (Building Management Systems Specialist for the California Academy of Science). Personal interview with Ari Harding, june 2009.
Susan Wels, California Academy of Sciences: Architecture in Harmony with Nature. 2008.