Architecture or ? Case Study of the Roof and Skylight Replacement on the Everson of Art, Syracuse, NY

Dennis C. Spina, AIA Bell & Spina, P.C. Architects ­ Planners Syracuse, New York

Roof Consultants Institute

Proceeedings of the RCI 21st International Convention Baskaran and Ko ­ 139 ABSTRACT

The Everson Museum of Art was designed by I.M. Pei in 1965 and occupied in 1968. When it opened it was (and still is) considered a work of art in itself. It is perceived as a piece of sculpture placed on a podium that one walks around. Indeed, one of the comments often heard is that it’s difficult to find the entrance – almost forcing visitors to walk around it to experience the structure. ( Architectural Forum ­ June 1969.)

By 2002, the museum had a long history of water infiltration. Bell & Spina was retained to design the replacement of the roofing and skylight systems. What we thought would be a rather straightforward project turned into a two­ year study to “do the right thing” for this landmark building, the museum, and its collection. This is a case study of the technical exploration to solve a number of waterproofing issues, along with the aesthetic concerns that pre­ sented themselves. It is also a study of the process and the discussions of the various strongly­held points of view, and of the compromises made to solve both aesthetics and waterproofing issues.

SPEAKER DENNIS SPINA is principal in charge of the project and senior roof consultant for Bell & Spina. He has spent a year with the building committee exploring options and facilitating the discussion on what the right thing is for the museum, its collection, its mission and the long­term integrity of the building's structure and the desire to retain the original design esthetic.

Spina ­ 140 Proceeedings of the RCI 21st International Convention Architecture or Sculpture? Case Study of the Roof and Skylight Replacement on the Everson Museum of Art, Syracuse, NY

INTRODUCTION The Everson Museum of Art was designed by I.M. Pei in 1965 and occupied in 1968. When it opened it was (and still is) consid­ ered a work of art in itself. It is perceived as a piece of sculpture placed on a podium that one walks around. Indeed, one of the comments often heard is that it is difficult to find the entrance – almost forcing visitors to walk around it to experience the struc­ ture. (Architectural Forum – June 1969.) By 2002, the museum had a long history of water infiltration. Bell & Spina was retained to de­ sign the replacement of the roof­ ing and skylight systems. What we thought would be a rather straightforward project turned into a two­year study to “do the right thing” for this landmark building, the museum, and its collection. This is a case study of the technical exploration to solve a number of waterproofing issues, along with the aesthetic concerns that presented themselves. It is also a study of the process and the discussions of the various strongly­held points of view, and of the compromises made to solve Photos by Ezra Stoller. both aesthetics and waterproofing issues. SETTING THE STAGE the administration wing. Each The skylights are divided by presented a different design chal­ mullions at 40 inches on center.

Existing Conditions lenge. This skylight mullion grid match­

es the building's structural grid, There are ten separate roof The existing roof membrane which is designed on a ten­foot areas, four galleries, a central was a fully adhered EPDM, which module subdivided into 40­inch sculpture court, an administra­ was installed over tapered insula­ increments. tive wing, the auditorium, sepa­ tion on top of the original coal tar, rate stair tower, and two canopies. built­up membrane, which was All of the building's HVAC air There are five skylights – two at installed directly to a reinforced intakes and exhausts are at roof the sculpture court and three on concrete structural deck. level.

Proceeedings of the RCI 21st International Convention Spina ­ 141 The Building Committee The museum had established a building committee of 14 people. Like a typical committee, the make­up consisted of representa­ tives from the museum, the muse­ um's board, and the community. It included the museum's direc­ tor, finance and facilities staff, six architects, two engineers, a con­ tractor, a developer, and the chairman: • Mark Robbins, dean of the School of Architecture at • Toby Nadel, AIA, a roof consultant • Manny Barbas, AIA, The In the initial meeting, October Existing configurations and deputy commissioner of 30, 2002, the building committee details of the skylights match the facilities for Onondaga directed that we make the build­ shop drawings. County ing watertight while respecting the original design intent. We also Design Conditions • Cal Bowne, AIA, principal consciously imposed constraints of the firm that was the Weather conditions in upstate upon ourselves not to impact the associated architect dur­ New York can be extreme. It original design. ing the original construc­ snows a lot. Last year (season), Syracuse received 181 inches of tion. Design Intent snow, which was not a record. It • David Nutting, AIA, princi­ The first thing we did was to is also cold, with a low design pal of VIP Structures, a review the existing conditions and temperature of minus 10 degrees large design/build firm. original construction documents. F in winter; and for the summer, The integration of the mechanical 90 degrees F, for a delta T of 100 • Bob Haley, AIA, Ashley­ systems with the structure is degrees. The freeze/thaw cycles McGraw Architects impressive, and the systems are are many over the course of a sin­ • Ravi Raman, PE, an elec­ well hidden. The detailing on the gle season. The American Associ­ trical engineer and princi­ skylights is also unique. To keep ation of ' recommenda­ pal of Ram Tech Engi­ the pure geometric form of the tion for interior climate conditions neers, a large engineering design, the skylights have mini­ is 75 degrees F and 50% relative firm. mal slope. For example, skylights humidity. These interior condi­ over the sculpture court are tions are set to protect paintings

• Edgar Galson PE, a mech­ detailed at a slope of 1/2 inch and are a requirement in order to

anical engineer and the re­ over a distance of 6'­8", or about receive some traveling exhibits.

tired principal of Galson 1/16" per foot. Detailing on the

Engineers, a large engi­ original contract documents indi­ PROCESS neering firm. cate no mullions were intended. Our analysis of existing condi­ • Jim Taylor, president of The triangular skylights in the tions found challenges that J.D. Taylor Construction administration wing (skylight #3) revolved around the lack of height Co. are the only skylights without to adequately flash and counter mullions and appear to closely flash the roof membrane and the • Gary Pickard, president of conform to the contract docu­ skylights. The single glazed sky­ a development company ments. With the skylights having lights had the potential for con­ such minimal slope, the caps • Ed Kochain, chairman of densation. All of the galleries and block the flow of water. The sky­ the committee and also the sculpture court lacked over­ light glass would have to be the deputy county execu­ flow protection. It was obvious cleaned five to six times a year, as tive of Onondaga County. that the key design element was dirt and algae would accumulate.

Spina ­ 142 Proceeedings of the RCI 21st International Convention Point load glazing examples. the skylights. Solve those issues found that the heat transfer CONSTRAINTS LIFTED and the rest should fall into place. through the concrete mass would Our office expressed confi­ keep the frame system from con­ We started to brainstorm dence in the design while sharing densating. We also found numer­ ideas to provide the original the concerns of the committee. We ous thermal shorts at the galleries design intent ­ i.e., a transparent, also stated that given the con­ that were outside our project weather­tight separation between straints it was a viable solution. scope to correct as it would re­ galleries. Our research lead us to The committee then decided to quire renovating the interior. Our contacting several glass manufac­ give us latitude to solve the design $1.2 million budget was approved turers. We settled on a point load problem without any constraints. and we proceeded to design devel­ glazing system for the following opment. Back to the drawing board. We reasons: came up with two possible solu­ A week after we presented the • Insulated laminated glass tions: design development (DD) report, was able to be provided. there was another meeting at Option 1 • Provided allowance for which the committee voiced its movement. (Glass moves concerns over a system that H a y d e n independent of frame sys­ would be custom designed for the Planetarium in tem.) museum by a firm from outside New York City the country. Their concerns gave us an idea • Proven track record. included that the museum may that on the sur­

• Single point responsibility. have difficulty getting timely response for • Ten­year watertight war­ replacement of broken rantee. glass or repair to any

leaks or to problems Schematic design proceeded with the system. The with concept sketches and analy­ primary concern was sis. We applied for and were a great hesitancy to awarded a grant through SATOP, rely on gaskets and the NASA­sponsored “Space Alli­ sealant as a primary ance Technology Outreach Pro­ watertight compo­ gram,” to enlist the help of Syra­ nent. cuse University to determine the potential for condensation of the point­loaded frame system and glass under design conditions. We Space elevations.

Proceeedings of the RCI 21st International Convention Spina ­ 143 face may seem extreme, but when and would make the building out additional renovations analyzed, is actually a good solu­ more visible at night. It also had and even then the humid­ tion. Placing a glass structure no pretense of being part of the ity level may have to mod­ around the Everson would give original design. However, the ulate with the exterior the museum its original design existing skylight over the direc­ temperature so as not to intent of a clear void between gal­ tor's office is not visible from the condensate), limiting the leries, the bonus of additional exterior. Placing a clerestory at shows that the museum exhibit space, and a strong fresh the director's office would not can exhibit. visual presence. After all, why not extenuate an original design ele­ • The skylights have leaked put a “sculpture” in its own muse­ ment, but add one. The proposal, um? therefore, was to eliminate that since the building opened.

skylight and create a faux skylight • The concerns identified At $25 to $30 million, the inside with lighting. committee passed on this solu­ need to be addressed, or conditions will continue to tion. Half of the committee liked the concept; the other half were con­ be detrimental to the art and the building. Option 2 cerned that it was too much of a departure. Our second exploration in­ All the above led to the discus­ volved the elimination of all hori­ sion that the museum is not able SUCCESS REDEFINED zontal glazing by turning the sky­ to fulfill its mission if the building lights into clerestories. Placement The “Mission” of any museum cannot be made watertight. of clerestories extenuated the is to: “house, protect, preserve, Criteria established by the muse­ original design. The concept and present to the current and um building committee, in order focused attention on the entrance future generations.” of priority: The discus­ 1. Watertight integrity of the sion was wide­ building.

ranging, but 2. Concern of aesthetics. boiled down to the fact that: Other criteria: • I n t e r i o r 1. Utilize a “standard” manu­ d e s i g n factured skylight system.

conditions cannot be 2. Installer should be local, within a 100­mile radius, o b t a i n e d (i.e., hu­ with the ability to provide and install replacement midity of 50% with­ glass.

Standard Manufactured Skylights Option 2 concept. A “standard” skylight from most manufacturers has a mini­ mum slope of 17 to 18 degrees. Using the minimal slope, we established some heights and pre­ sented the results, which our office and all members of the com­ mittee felt were not acceptable. One manufacturer (Naturalite) offers a low­pitch skylight with a minimum slope requirement of 1/2­inch per foot. This system also could be integrated with a curtain wall system. Computer­ generated elevations indicating

Spina ­ 144 Proceeedings of the RCI 21st International Convention the visual impact were pre­ sented. All thought the visu­ al impact was acceptable. This solution was proposed to and approved by the State Historic Preservation Office (SHPO) as required by one of the grant sources. With the skylight manu­ facturer selected, we pro­ ceeded to tackle the details of integrating the roof sys­ tem. DESIGN Exploring options: low pitch skylight.

Asbestos Abatement Membrane court roof was available from The residual asphalt mastic The ten roof areas were each gallery “C.” Gallery “D” was on the parapets and the insula­ small enough to typically be made accessed by walking across the tion on some of the roof drainage watertight in a day. However, the skylight or placing a sheet of ply­ piping had to be abated. lightweight system used to pro­ wood across the skylight. Access vide slope to drain would have to Gallery “B” was by plywood over Drains residual moisture. Therefore, a the skylight and then by ladder. PVC membrane using the 2001 The location of the existing We designed fixed ladders that System that allows drying out drains did not present a logical swing into position to provide over time was chosen. The system solution using tapered insulation access across the skylights. All also had the advantage of hurri­ board. We proposed to use an other roof sections, stair, auditori­ cane wind ratings of 120 mph – insulated, lightweight concrete um, administration wing, and extra insurance for the museum, system to provide slope to drain. canopies are available only by lad­ even though winds of that speed This system also allowed us to der from the ground. are not experienced in the area. In abandon the four scupper drains on the center sculpture court.

Debris plugs scupper drains more easily than standard Below: cracks. roof drains. Indeed, on one site investigation visit we found six inches of standing water on the sculpture court roof. The need for overflow protection was apparent. The gallery parapets varied in height from 16 to 60 inches. Overflow protection on the galleries was provided by core drilling through the walls, and on the sculpture court, with overflow drains. Scupper drains on the canopies were retrofitted.

Access There are two roof access hatches – one in gallery “B” and one in gallery “C.” A lad­ der to the center sculpture

Proceeedings of the RCI 21st International Convention Spiina ­ 145 the future, when the membrane needs to be replaced, other options for a membrane will be available to be placed over the dry, lightweight fill.

Skylights 1 and 2 There were no real waterproofing compromises made in the detailing of the sculpture court skylights. Minimum flashing heights and counter flashing were able to be achieved and allowances for movement pro­ vided. All the above sailed through the committee. Committee membership changed and the new mem­ bers started asking the same questions we had asked ourselves months earlier. Answering their concerns resulted in the committee revisiting the following issues.

Parapet/Edge Detail There was much debate over the need to cover the building's concrete parapet coping and the visual impact it would create. Multiple options were explored, such as use of a coating and cutting a reglet into the top of the parapet. In addition, five coping samples were made in stainless steel, freedom grey, black, and bronze. These samples were then placed on the upper and lower parapets and viewed by the committee mem­ bers from street level. Surprisingly, all the samples placed on the upper sections appeared the same and disappeared against the overcast sky. In the end, the committee voted. There were three edge details proposed: coping mock­up #1, which did not break the edge and relied on sealant for protection; coping mock­up #5, which had been previously approved; and a variation called 5A that had a straight rather than 45­degree drip, intended to present less of a shadow line. Each committee member voted and explained his reasoning: • We must cover the edge ­ 5A

Spina ­ 146 Proceeedings of the RCI 21st International Convention • Agree likes ­ 5A Total vote: six for detail #5A; element and voted to keep this

two for detail #1, one for detail #5. skylight in its current configura­ • Coping #1 needs mainte­ tion with the recognition that it nance, therefore 5A Administration Wing Skylights will be below the manufacturer's minimum slope requirements and • The criteria established by Options for the disposition of would not be guaranteed water­ the commitee was water­ Skylights #3, 4, and 5 were previ­ tight. tighness first; the second ously agreed to and followed the

was aesthetics. Coping #1 clerestory proposal option 2. flips the order of that cri­ Skylight 5 teria; therefore ­ 5A. Skylight 3 This skylight was similar to skylight 4, with a lack of slope • The drip edge is a better The issue with this skylight and height to flash. It had previ­ solution ­ 5 was that the curb had to be raised ously been agreed that the direc­ to accommodate the lightweight • There is a need to inspect tor's office skylight would be elim­ fill and to properly counter flash every year; therefore, I inated. The committee desired to the membrane. The glass had to suggest a compromise of keep this design element and be an insulating type. Keeping the the lower sections using voted to keep it with the recogni­ existing glazing slot location detail #1 and in the upper tion that it would be below the would require enlargement and a gallery sections, use 5 or manufacturer's minimum slope reduction of the light well width to 5A. requirements and would not be accommodate sill to roof counter­ guaranteed watertight. • Any metal added to the flashing. Raising the skylight and building is a detriment to keeping the same slope would FINAL THOUGHTS the aesthetics of this result in the peak of the triangu­ building. I do not believe lar skylight being above the con­ Next month it will be three any coping cap is required crete fins. In the end, we raised and a half years since we started on the upper gallery roofs. the skylight, lowered the slope, this project. What have we The lower areas need a and provided insulated glazing in learned that we can pass on? cap but should be zero a frame system. • In order to have water run visibility #1 or earlier pro­ downhill, you must pro­ posed modified detail with Skylight 4 vide the hill. (Positive a reglet. The difficulty here was lack of slope is a requirement.)

• Prefers nothing but is of slope and height to flash. Also, • Do not rely on sealant as the opinion that we will this skylight had a vertical glazing your primary line of de­ lose the lower parapet component that was installed in a fense against water infil­ walls; therefore, detail 5A. slot in the concrete. The glass had to be installed from the top. At tration. (You should flash

• Would like to see nothing only 20 inches wide, it did not and counter flash.) on the walls but does not make sense to make it a triangu­ • Involvement of your client want to rely on fool­proof lar skylight. The committee is a good thing. (This com­ maintenance; therefore, 5A. desired not to change this design mittee was engaged and

Proceeedings of the RCI 21st International Convention Spina ­ 147 helped define success.) • There is no such thing as only one solution. (We explored many solutions.) • Beauty is in the eye of the beholder. (Solutions that bothered some aestheti­ cally were embraced by others.) In the end, we collectively endeavored to preserve this land­ mark building, the museum's mission, and its art collection.

Skylight 5 was completely covered over. The light well became more gallery space.

Spina ­ 148 Proceeedings of the RCI 21st International Convention