What Made Lucy Rot? A Case Study of Cyclical Moisture Absorption Margaret Westfield and Richard I. Ortega, with Ernest A. Conrad

The New Jersey shore, especially Absecon Island and Atlantic City, provided recreation for citizens of the Philadelphia area. As that economy was especially strong during this period, real estate speculators worked feverishly to attract prospective developers and home builders to the shore area.

James Vincent de Paul Lafferty, Jr. (1856-98), engineer, inventor, and real estate speculator, owned a tract of beachfront on the southeast portion of Absecon Island south of Atlantic City, but he needed an attraction, something novel, to draw prospective buyers. He decided to erect a building in the shape of an elephant. Designed by William Free, a Philadelphia architect, and built in 1881 at a cost of $38,000, Lucy, as the building was subsequently named, attracted visitors who came to marvel at her sheer scale. Visitors would be invited to take a tour, to see the panoramic views from the observation deck or howdah on Lucy's back, and to see the tracts of land Lafferty had available for sale.

Lafferty built two other elephant structures. One of these, known as the Light of Asia, was constructed in South Cape May, New Jersey, in 1884 but was never maintained; it was torn down c. 1900. The other was built at Coney Island, New York, also in 1884. This 122-foot-tall structure, appropriately named Elephantine Colossus, had 31 rooms; it was destroyed by fire in 1896. Lucy is the only survivor.

In 1887 Lafferty sold Lucy to Anthony Gertzen, along with other parcels of land. Gertzen purchased the Turkish Pavilion from the 1876 Centennial Exhibition in Philadelphia and reconstructed it behind Lucy as the Elephant Hotel. The Gertzen family ran Lucy as a tourist attraction for nearly eighty years, offering ten-cent tours.

By the late 1960s, Lucy had weathered considerably, was in serious disrepair, and was in the way of new real estate development. In response to her threatened demolition, the Save Lucy Committee was formed, and Lucy was donated to the City of Margate in 1970 with the stipulation that she be moved off-site within thirty days. Fundraising began immediately to relocate the 90-ton building to a nearby city park. The move to the present location at 9200 Atlantic Avenue occurred on July 20, 1970. An extensive stabilization and restoration program started soon thereafter.

As funding became available, additional phases of the restoration were effected. Improvements were made in deference to modern code requirements and programmatic needs. Both a Halon fire-suppression system and a dry-pipe sprinkler system were installed. Electric service and wiring were modernized. In anticipation of the building's use as a museum, heating and air conditioning were installed in 1984. To enhance the thermal performance of the building envelope, fiberglass batt insulation was installed between the curved framing elements against the sheathing, with the foil side facing the interior. However, the partitions designed to reconstruct the historic appearance of the interior, create an envelope for the HVAC system, and conceal the insulation were never built.

In 1987, New Jersey voters approved a referendum for bonds to provide matching funds for the preservation of significant historic properties controlled by public agencies and non-profit organizations. The Save Lucy Committee sought this funding to complete the final, interior, phase of Lucy's restoration. In 1992 Margaret Westfield, R.A., of Westfield Architects & Preservation Consultants, was engaged to prepare the documents for the interior restoration. She determined that significant damage to the building fabric had occurred in the twenty years since the initial restoration. It was clear that this damage would have to be addressed before the interior restoration could proceed. She engaged Richard I. Ortega, P.E., R.A., as a consultant to address structural and building conservation issues. Ortega discovered extensive damage to previously restored elements of the building, including the exterior sheathing and secondary framing members. Based on these findings, the scope of work had to be changed completely.

The initial survey uncovered the following:

• Inspection of the exposed, secondary framing elements (i.e., the framing elements that form the body and skin) revealed extensive evidence of water saturation, staining, rot, and fruiting bodies from the rot.
• Both the insulation and the wood sheathing below it were saturated with water.
• Probing with a scratch awl revealed that the exterior surfaces of many of the curved ribs that shape the body were soft with rot.
• There was a pattern of rotted sheathing in the belly of the elephant, directly below the steel trusses and the iron piping for the sprinkler system.
• Much more rust was discovered on the bottom chords of the trusses, especially the bottom surfaces of the bottom chords, than anywhere else on the trusses, despite the fact that the trusses are substantially inboard from the exterior of the building. There was a similar rusting pattern on the iron piping.
• In low areas of the body (the chin and belly) there were rot, fruiting bodies, and evidence of water ponding but no indication of a nearby point of water entry.
• In some areas the intumescent (fire-retardant) paint applied to the interior of the wood structure had more structural integrity than the wood sheathing; when the paint film was lifted, the rotted sheathing also pulled away, revealing the interior surface of the exterior metal skin.
• On the exterior there was a notable amount of rust and water dripping from the bottom tips of Lucy's ears. The seams of the exterior metal pans were rusting, and the paint finish on the red blanket on Lucy's back was discoloring the gray body below it.

After interviewing Josephine Harron, president of the Save Lucy Committee since 1970, it became clear that there had been a history of water-related problems with the skin of the building since the completion of the exterior restoration. For each problem a cause had been identified, and what was expected to be a solution effected:

The paint failed to adhere properly to the metal skin. The problem was attributed to the paint, and the manufacturer prepared a new formulation that the contractor applied at no charge.

There were isolated areas of interior sheathing dampness. The problem was identified as rainwater entering through improperly folded seams in the metal pans of the exterior skin. Strips of rubber were applied over the seams in the upper half of the structure to seal them. This solution apparently rectified the problem but had a negative visual impact.

It was apparent in 1992 that there was very active deterioration caused by water, but the source was not readily apparent. The history of water problems and remedies did not conform to the apparent pattern of deterioration. The evidence suggested that the building was suffering from high humidity and condensation problems.

Fortunately, there was no damage to the primary structural system. However, because of the apparent extent of the damage to the skin and the potential for losing large areas of it as the sheathing and secondary framing rotted, it was decided that the interior restoration should be postponed and a detailed survey done to verify the extent of the damage.

A contractor removed all the insulation and exposed the sheathing and secondary framing. The entire interior surface of the building skin was then surveyed, and the areas of rot were annotated on the interior elevations. It became clear that there was some rot damage to all areas covered with insulation. However, the most severe damage was concentrated in the bottom half of the building. Except for very localized damage below the corners of windows, there was no correlation between the rot and potential sources of exterior water entry. Indeed, logical places for water entry at the top of the building (such as where the howdah supports penetrated the skin to bear on the trusses or around the skylight) revealed almost no evidence of water entry.

The survey appeared to confirm that the damage was the result of condensation, but it was not clear what processes were causing the problems, and caution was exercised not to effect repairs and alterations without first understanding how Lucy was rotting. It was decided not to reinstall the insulation, as it only seemed to exacerbate the problem and not to run the air-conditioning equipment until its impact could be determined.

Even though the physical evidence suggested that some form of microclimate was causing rapid cycles of wetting and drying through condensation, it was not known when it occurred and under what conditions. Clearly, a monitoring program seemed warranted. Ernest Conrad, P.E. was selected to oversee the monitoring project and serve as the project's mechanical consultant.

In normal building configurations, the day/night cycle of moisture absorption and desorption within the fabric would be at a net equilibrium daily between the effects of daytime infiltration dilution and nighttime building mass absorption. Usually, the outside air that infiltrates the building's interior is dryer than the inside air and absorbs moisture as it passes through. This moisture is driven from the interior woodwork by the effects of solar heating. Thus, equilibrium is generally a daily cycle where the moisture vapor added to the interior air from the desorption of water from wood is removed by infiltrating outdoor air. With Lucy, equilibrium does not occur and thus saturation results. Simply put, the solar heating effects on the building have about six hours each day to drive out water that has been absorbed from cooler, humid air during the previous 18 hours. Analyses by Conrad indicated that for equilibrium to be reached, natural ventilation alone could not exhaust enough dilution air to maintain an acceptable RH equilibrium or permit tolerable occupancy inside Lucy. Calculations disclosed that, even with forced ventilation through the existing vents in Lucy's belly, room conditions would routinely approach 100F in the summer, with a RH in the 60% to 80% range.

It is clear from the analysis of the monitoring data that this imbalance is a product of a flawed building design that has existed from the day Lucy was built. Clad in a non-breathing metal skin with minimal openings for ventilation in a relatively humid environment, Lucy could best be compared to a bottle terrarium, except that in Lucy's case there was a constant source of humid outside air entering the building and exacerbating the damaging processes.
Thus, Lucy is an exaggerated example of the remarkably powerful effects that moisture can have on a building and how easily the symptoms can be masked or misdiagnosed. Previous measures to eliminate the sources of the water problems focused on perceived failures of the exterior envelope to shed water, not on the building as a micro-climate being unable to achieve equilibrium with the surrounding environment. What had historically been attributed to poor workmanship and/or product failure, may, in fact, have been caused by the terrarium effect.

In retrospect, it is clear that Lucy has always rotted from the inside out. Reexamination of the 1973 pre-restoration photograph (Fig. 4), shows that the pattern of damage so severe in the belly, low-lying body elements, and the flanks is remarkably similar to the pattern found in the 1992 survey of the rotted areas. The restoration architects, in 1973, did not have the benefit of the physical evidence now available. The building had been neglected for so many years that the massive deterioration was logically attributed to breakdowns in the building envelope.

The structural engineer who designed the steel trusses in 1973 confirmed the original wood trusses had rotted. The steel trusses did not change the dynamics of the problems with Lucy, only where and how they manifested themselves. Where the timber trusses would have absorbed the excess moisture in a constant effort to reach an equilibrium moisture content relative to the ambient conditions (and then rotted), the steel trusses provided an element on which condensation formed and then dripped on adjacent wood materials. The sprinkler piping provided a similar opportunity for condensation to form.

The introduction of air-conditioning operated for cooling rather than dehumidification, the use of set-back thermostats, and, worst of all, the installation of foil-backed insulation against the sheathing proved very damaging. All these elements were introduced as standard energy conservation measures (bear in mind the work was done in 1984), but in this building they served only to accelerate the destructive processes. While it is clear that the sheathing never could have desorbed as much moisture as it absorbed in each diurnal cycle, it was also clear from the data that some areas, especially the upper portions that were in direct sunlight, had the moisture driven out of the wood daily by the heat of the sun. This moisture, of course, added to the overall interior RH and in large part was diffused throughout the building or reabsorbed by other wood elements. The foil-backed insulation trapped the moisture driven out of the wood sheathing by the sun. In most cases, this water appears to have condensed on the inside surface of the metal skin and flowed to lower areas where it puddled on the wood sheathing. The net effect was to trap and concentrate the moisture, in liquid form, in particular parts of the building.

The problems are now being addressed. The insulation has been removed, and the existing HVAC units have been reprogrammed primarily for dehumidification and only secondarily for cooling. In the long term, the interior partition walls must be reconstructed to create a conditioned, insulated box. This inner box will follow the walls, vaulted ceiling, and floor line of the interior main hall. The perimeter voids will be isolated for active exhaust ventilation at higher ambient temperatures. The existing belly vents, which were too small to provide effective ventilation, will be reused as intakes for two new mechanical heating and ventilating air handlers. Exfiltration passages will be created at all of Lucy's dead-end spaces, such as her ears, trunk, tongue, tail, and legs. These adjustments should finally resolve the moisture problems that have plagued Lucy since her construction. A long-term monitoring capability will be installed to provide feedback for future performance evaluations.

Because Lucy is a unique structure, it is difficult to generalize these findings for application to other buildings. There are, however, lessons to be learned from this example.

First, the obvious solution may not be right. Other possible scenarios that conform to the evidence, regardless of how improbable they may at first seem, must be explored.

Second, while not every building is as unusual as Lucy, every building is unique. Therefore, standard design solutions cannot be applied to the apparent problems without first exploring the potential negative ramifications.

Third, the importance of monitoring cannot be overestimated. Monitoring provides the factual evidence to corroborate observations and substantiate theoretical conclusions.

Investigating Lucy's problems was a fascinating exercise in which a multidisciplinary team was able to apply a variety of complementary skills. In addition to the individuals mentioned above, Morgan Phillips of Phillips Architectural Conservation consulted on the potential consolidation of the damaged sheathing; Lorraine Schnabel of John Milner Associates developed a replication mix to match the original plaster; Frank S. Welsh, architectural coatings consultant, provided finishes analyses; architectural historian Rebecca A. Hunt undertook extensive documentary research; and the staff of the New Jersey Historic Trust and the New Jersey Historic Preservation Office provided guidance at the state level.

Margaret Westfield, R.A. is a historic architect and partner in the firm of Westfield Architects & Preservation Consultants of Haddon Heights, New Jersey.
Richard I. Ortega, P.E., R.A. is principal of Ortega Consulting of Media, Pennsylvania, a firm specializing in addressing the structural and conservation needs of historic buildings.
Ernest A. Conrad, P.E. is a mechanical engineer and serves as principal of Landmark Facilities Group, Inc. of East Norwalk, Connecticut.