Investigation into Water Leaks at the Underground Art gallery

Investigation into Water Leaks at the Underground Art Gallery of Cape Cod


In February of 2011 the Underground Art Gallery experienced water leaking into the display/studio space from more than one spot where the roof and south wall come together. The gallery, built in 1988 by the late architect Malcolm Wells, was designed to be a space for his wife’s art and as his own architecture office as well as an example of what Wells termed a “gentle architecture”. Leaks, however, had occurred before as Mr. Wells very honestly spoke of them in more than one of his books. The cutting edge designer had included an experimental “knife-edge shelf” on the south side which cantilevered earth beyond the glazing on the south of the building (see figure 01).

The shelf had been calculated to create summer shade while allowing earth to almost “float” as opposed to the more traditional massive structure found in similar situations, or a shingled parapet seen on so many underground homes of the past. On page 59 of How to Build an Underground House Wells discusses this aesthetic consideration complete with an illustration showing a bird expressing disgust with the “fat” overhang of a traditional fascia.

Aesthetics were probably not on Mrs. Wells’ mind when the leaks began again this year. The late architect’s widow had difficulty finding someone knowledgeable enough about underground building to deal with the leaky situation. Having been enormously influenced by Mr. Wells as well as having built my own underground home in Maine, I happily agreed to try to help upon being contacted by Mrs. Wells. She remembered me from my correspondence with her husband as well as having had the opportunity to meet them both at the Underground Art Gallery in 2003.

On my July 2011 return to Cape Cod, some eight years after my initial visit, I found attempts had already been made to remedy the situation by placing more bituthene at the suspected sources of the problem identified in books and notes by Mr. Wells (or Mac as he was known to family and friends). Of course this work first involved removing all of the earth from the experimental knife-edge shelf. Interestingly, the water entering the building was slowed considerably but not stopped by these dutiful interventions.

More work would be needed. Investigation of the roof in general found up to two horizontal inches (and more at corners) of bituthene on the west, north, and east sides of the building to be shy of some earth cover where it wore away presumably by water and wind erosion over the last twenty three years ultimately allowing for UV degradation (see figure 02).

Stucco at the southeast and southwest corners showed significant compromise as well, allowing water access to the wood framing beneath it, thereby allowing termites to access to moist wood and polystyrene insulation. Regrettably, the stucco detail was poorly executed during initial construction relying on “gooped-in” bituthene (petroleum based waterproofing) to establish water tightness. Much of the south wall has been found to be in a disastrous state with entire wooden (non pressure treated) studs and top and bottom plates having been consumed by termites and/or rot. Fortunately, the south wall bears no structural weight of the roof (steel reinforced concrete on steel deck) that cantilevers past supporting “tree trunk posts” some five feet to the interior of the wall.

Weighing heavily on our minds however, was the water that still entered the building despite repairs at the suspected leaks. More earth removal from the south side of the roof would be needed in order to eventually “patch in” a full sheet of bituthene membrane east to west from the surface of the roof down over the vertical aspect of the knife-edge, and onto the actual “shelf”. By dong this water would not be able to backup underneath the membrane in the event of an ice dam, as Mac had written about. Placed over the new bituthene would be drainage matt virtually guaranteeing that water coming through the earth would then quickly flow off the shelf. One-inch insulation might also be added to the vertical aspect of the steel shelf in order to retard heat loss.

Instead, it was decided earth would not be put back on the shelf as it had been experimental from the beginning and the experiment did not fare well in terms of water tightness, though aesthetically it was lovely indeed. The shelf is actually six ten-foot L-shaped sections that were not welded together where they joined (speculatively to prevent them from acting as one unit in expansion and contraction?) and it was at these joints that water has certainly been gaining access. It may have been that the bituthene membrane covering this joint could not bridge the gap created by contracting units. Upon completion of the restoration the knife-edge shelf will be covered with plant vines growing on a trellis like renovation. This will allow one to keep an eye in future years on the membrane where it joins the knife-edge shelf.

On inspecting the north edge of the roof a copper flashing was found, initially added to make up for the roof that ended shy of the wood trim around the windows of the north wall. This flashing had been tacked onto the vertical aspect of a narrow fascia board rather than coming off of the roof horizontally first and followed with a membrane to cover the joint between flashing and roofing. There was no evidence of leaking on the north side. However the roof clearly needed to be extended to a proper distance in order to allow water to drip past the wall beneath, and as already described exposed bituthene needed to be restored.
Also in need of repair were the east and west ends of the roof that relied on a precarious detail in which a pressure treated 2X4s ran along the wall serving to extend the roofs bituthene waterproofing out over the exterior polystyrene insulation of the walls meant to keep the concrete mass as part of the living space (while bituthene does also run off the roof and directly onto the concrete wall). Unfortunately, the stucco-covered insulation was proud of the 2X4 roof detail in several places allowing water to gain access underneath the stucco of the walls. In light of the above issues it was decided that earth would be removed from all four sides of the roof to allow for retrofitting of the problems as stated.

As digging continued roots up to ¾ of an inch thick (from several trees that were growing on the roof – now removed) were found to have migrated everywhere. Not only did roots burrow their way under, over, and around insulation (having been placed over the waterproofing membrane and without a root barrier – not known to be needed in 1988), but through it as well, often separating sheets of polystyrene and allowing earth to wash in-between and underneath it thereby lifting and moving it where the largest roots created the biggest voids. Most interestingly, each sheet of insulation taken off the roof was impossibly heavy with water to the point that they broke under their own weight and the pieces still quite heavy for one person to carry alone.

Underneath the insulation, the bituthene waterproofing membrane was found to be intact everywhere it was visualized (and had been protected from UV). In fact it showed no wear or other sign of concern other than having been “rippled” which may have been caused by roots pushing against it as well as insulation moving slightly due to root invasion. This meant that the initial impression of the insulation and earth slowly sliding down the seven percent roof grade was happily wrong. In a few places very small fibrous roots were found to have “integrated” themselves with the edges of bituthene but not growing in underneath it. The approximately two-inch overlap of one bituthene layer to the next was found to be intact where visualized. The bituthene looks as though it should not only hold up for another twenty-three years but much longer indeed.

During the lengthy process of excavating all the earth, the subject of starting from scratch with new insulation and the latest in green roof technology was discussed at length on numerous occasions. Ultimately, the decision to stay with the more conservative repair approach was selected each time for three reasons, the first being financial. Despite efforts at fundraising only slightly more than one thousand United States dollars has been raised from a handful of generous donations to date. Secondly, it is believed that water is not entering from anywhere yet uncovered as when the initial leaks are covered with plastic no water enters the building, and the bituthene uncovered looks to be in great shape. Lastly, it was felt that this was the approach Mac would have used in an effort to be most economical.

An economy for the respect of Nature was the foundation on which Mr. Wells based his philosophy. Today this is called conservation. The architect saw a future in which people worked with the environment as he realized early on the consequences of a society based on waste and greed. Mac was very generous with other people, but for himself he was always frugal. When he built the Underground Art Gallery he had several different people work on it at different times, and like most truly intellectual visionaries Mac may have been looking at the big picture more than at the minute details. Mac’s ideas are no less relevant due to these unforeseen circumstances. The designer’s ideas of an architecture that honors Nature are more important then ever. Indeed, earth sheltering is essential to a sustainable future. The green roof technology developed in recent years coupled with the lessons learned on this project allows for a bright earth sheltered future indeed.

It is the future for which our team has undertaken this restoration of no small feat. Primarily long time Underground Art Gallery employee Jenna (who describes herself as totally committed to Mac’s ideas as well as the current gallery) and myself have volunteered our time for this project. A project in which we have moved some 30 plus cubic yards of earth from the roof by hand, carefully removed insulation, as well as completing the relatively significant carpentry involved in extending the roof on three sides (see figures 03 and 04).


Mrs. Wells has helped on numerous occasions neglecting her need to run the gallery from which she makes a living, and struggling to find time to actually paint. Jenna has also brought willing volunteers, friends and family, to help dig, clear brush, and more. A professional steel worker, Ray Parks of RSP Welding in Brewster was hired to do needed retrofitting of the knife-edge shelf. Brendan Shea from Recover Green Roofs, LLC of Somerville, Massachusetts has been helping with the plans for the roof repair and will do the actual membrane work for cost. Mr. Shea, like so many others, was greatly influenced by Mac and wants to help preserve the legacy of Wells’ architecture. Underground Art Gallery picture framer Dave designed the layout adorning canvas bags offered for donations of one hundred dollars or more (with bags being sent retroactively to applicable donations to date due to bag printing being delayed several times). Please look for the image that Dave hid in the picture as his tribute to Mac! Another source of help has been Organic architect James Schildroth of Maine who volunteered valuable knowledge by phone on several occasions. Recently it was found out that flashing work including drip edge as well as “boots” fabricated to fit over the junction between shelf plates (where the leaks occurred) as well as covering exposed wood at the south corners will cost nearly 3000.00 USD. Most recently a contract was signed to rebuild the chimney above the roofline costing more than 2000.00 USD.

Even on completion of the roof repairs, volunteers will be needed for work involving the south wall that still needs to be almost entirely replaced! The team hopes to reuse the sliding glass door units as well as the glass of the transom windows above. If possible the wall work will be completed this winter, as that is the galleries off-season. Come spring stone retaining walls will need to be built at each corner of the gallery, insulation fitted back to the roof and walls, stucco reapplied, and then there is all that earth in need of being returned to the roof! By summer, however, the gallery roof will have rebounded with newly growing grasses, myrtle, weeds, and wildflowers just as Mac would like . . .

Please visit the blog HYPERLINK "http://www.savetheroof.blogspot.com" www.savetheroof.blogspot.com where Mrs. Wells (with the aid of Ms. Wendy Mathias who has helped with all things computer related) frequently enters updates on the work and tells the story of the project and people involved from the beginning (some of whom are not named above but deserve due credit)! On the blog you will also find the mission statement of those involved in the restoration. Don’t miss the “older posts” available at the bottom of the page.

Toward a sustainable future,

Oliver Solmitz, BA-Arch, AAS, EMT-P
Brewster, Massachusetts
November 2011



Figure 01. View of the un-earthed knife-edge shelf looking from the east.

Figure 03. Excavation of the roof on the north side.

Figure 04. Carpentry to extend the roof.

Figure 02. UV damaged bituthene and water getting in behind stucco at southeast corner.