Last updated: December 3, 2025
Article
1992 Design Analysis of 1976 Reconstruction
NPS photo
Content adapted from the Bent’s Old Fort Maintenance Repair Plan prepared by Preservation Partnership, April 1992. This essay reflects the state of the reconstruction in 1992, when it was sixteen years old.
Bent's Old Fort, a large trapezoid-shaped adobe structure enclosing a central plaza was originally built around 1833. The Fort was reconstructed by the National Park Service in 1975-76 from historical documentation, drawings, and conjecture. The new fort began to require significant maintenance and repairs before the reconstruction was completed. Considerable redesign of the Fort construction methods and materials occurred during construction.
Problems with the reconstructed fort’s maintenance began with its adobe blocks and render as the structure was nearing completion. Additional complications related to the roofs, walls, wood structures, and metal conduit shortly followed. These issues have generally been addressed by the maintenance staff who have employed numerous techniques over the years with varying success. The ongoing need for maintenance stems from reconstruction design deficiencies and have defied simple solutions.
DESIGN INTENT
Reconstruction Background
The decision to reconstruct Bent's Old Fort was made in the 1960's by the National Park Service. The process leading to reconstruction was begun with historical research and archaeological excavation in 1963. Dwight Stinson, NPS Historian, completed his comprehensive history data section of the Historic Structure Report in 1966. This document, along with the archeological report, provided the research base for the preparation of construction documents.
Absence of Architectural Data Section
It is important to note that an architectural data section was never prepared for Bent's Old Fort. According to George Thorson, Project Architect, an NPS architect was originally scheduled to research and write this report, but when funding was eliminated, only the preliminary sketches of the fort were completed. This deficiency was not resolved, and a detailed architectural evaluation of the reconstruction proposal was never completed.
The design of the reconstruction was made without this vital architectural information and analysis. If the architectural data section had been completed, it is possible that excessive maintenance would not have been required and the unsuitability of natural adobe for the local climate would have been identified.
Design Philosophy
The Ken R. White Company of Denver was awarded the contract for preparation of the construction documents for the fort in July 1973. The overall philosophy of the reconstruction is described by George Thorson, Project Architect, as:
... the three challenges to research, design and construct a building that would be as authentic as possible in every detail and would provide visitors with a time-space trip back to 1846. At the same time, durability was a requirement in the construction of the fort—to make it last 100 years or more in contrast to the 16-year lifespan of the original fort. The twin requirements of authenticity and durability had seemingly inherent contradictions that could be solved only with creative ingenuity to achieve a delicate yet convincing design balance.
Resolution of Authenticity vs. Durability
This contradiction of authenticity and durability was not resolved. The "creative ingenuity" resulted in widespread incompatibility of materials and unconventional detailing. These design problems are further discussed below.
Reconstruction Data
While the archeological documentation provided clear evidence of the Fort's foundation, the reconstruction design relied upon limited documentation of the rest of the structure. Aside from the historic Abert sketches of 1846, the design was based largely on conjecture. There are no photographs of the fort, even from the later stagecoach period in the 1860's and 1870's. There was simply no way to verify the dimensions or proportions of the original fort.
According to the conclusions of the historical and archaeological reports, the fort originally measured approximately 140 ft. x 170 ft. It contained 29 lower rooms, 9 upper rooms, 2 circular bastions, and a 120'x 170' adobe corral abutting the main building. The adobes were 18"x 9"x 4" and the exterior walls varied from 2 to 3 adobes in thickness.
Abert’s sketches were described by the Project Architect as follows:
... four views of the interior of the plaza, and entrance elevation, and the overall roof plan of the fort.
The sketches are rough by nature and at best only partially accurate in terms of appearance, without size or scale.
The Project Architect described the percentage of conjecture involved in the reconstruction design as follows:
Although complicated by the divergent reports of eyewitnesses, some very obvious errors and fabrications, and, of course, the many hiatuses in the record, the historical and archeological data provided roughly thirty percent of the necessary material.
The Project Architect further stated that the archaeological report provided another thirty percent of the needed material for reconstruction. Without the research base of the architectural data section of the Historic Structure Report, the remaining forty percent of the design had to be entirely based on conjecture.
UNSUITABILITY OF ADOBE FOR THE LOCAL CLIMATE
Repairs Related to Climate
The unprotected adobe of the reconstructed Fort has proven to be incompatible with the local climate which is characterized by intense spring and summer rainstorms with driving winds. Severe erosion of the exterior natural adobe render and natural adobes forced a massive, nearly complete re-rendering effort prior to the dedication of the fort. The wind-driven rainfall was a major factor in the failure of the adobe render.
The north and west exterior walls of the fort have suffered the greatest degradation because most storms and prevailing winds travel this direction.
The collapse during the first two years of portions of the natural adobe exterior wythe can be attributed in part to erosion by weather.
The reconstruction design did not adequately prepare for local climatic factors, partly owing to the absence of an architectural data section in the 1966 Historic Structure Report.
Original "Santa Fe" Adobe Construction
The original circa 1833 fort was built in the Santa Fe tradition with flat mud roofs supported by log vigas and latias which spanned thick walls of sun-dried mud brick and mud render. The trading partners William and Charles Bent and Ceran St. Vrain built the fort along with a mercantile post in Taos to capitalize on the new trade route between Santa Fe and St. Louis, along the Mountain Branch of the Santa Fe Trail. The partners' choice of adobe for the construction of the fort was apparently based on familiarity with similar construction in New Mexico and the fact that the most experienced and available labor pool was from New Mexico.
Failure of the Original Adobe Construction
This flat-roofed adobe plaza was a building type without precedent in the Arkansas River Valley. While successful in Santa Fe, historical data suggests that the construction did not fare well in the wetter climate.
William Bent abandoned the fort fifteen years after construction and moved his trading operation to Big Timbers, 40 miles downstream along the Arkansas River. In this new location, Bent built a new fort of stone on high ground above the flood plain of the Arkansas. The reasons which motivated Bent to abandon the fort are undocumented, but undoubtedly were economic in nature. It is conjectured that the adobe building was simply falling apart after 15 years, and the maintenance had become an overwhelming and costly problem. Two reports from travelers in 1853, just 4 years after abandonment of the fort, describe its condition as "ruins" and "destroyed." A portion of the fort was later salvaged and used as a stagecoach station and post office into the 1870s.
If the 1834 construction had been successful, it is unlikely that Bent would have gone to the additional expense of building a new fort in 1853. Bent had apparently learned that
-
The flood plain of the Arkansas River was too risky for building, and;
-
The climate was too severe for unprotected adobe to survive without constant maintenance.
Examples of successful Adobe construction in the area.
Adobe as a building material is fairly successful in southeastern Colorado if it is protected with a slightly overhanging, pitched roof. Damage is usually confined to the splash zone at the base of the walls and maintenance is not unreasonable.
There are many good examples of adobe buildings not far from La Junta, which have survived from the 1860's and 1870's in good condition. The Baca House in Trinidad, the houses at Boggsville, a number of structures in Pinon Canyon and many homes and churches in the Purgatoire Valley are examples of early adobe construction modified from the Santa Fe model to adapt to the different climactic conditions found in Colorado. The Baca House, for example, is a two-story, hip-roofed structure with an exterior of adobe blocks and natural adobe render.
This render requires limited patching once or twice a year with a complete re-rendering every seven years.
It seems likely that the lack of a protective overhanging roof on the original construction of Bent's Old Fort was a major factor the failure of the adobe design.
NPS photo
INCOMPATIBILITY OF MATERIALS AND DESIGN DEFICIENCIES
In an effort to satisfy the often conflicting requirements of authenticity and durability, the reconstruction design of the fort's walls and roofs was a unique combination of traditional Santa Fe adobe and modern masonry design. These hybrid walls and roofs required, in turn, unique details for structural continuity, flashing, finishing, and drainage. It can be demonstrated that many of the fort's exhaustive maintenance efforts have been directed to problems which are due to an incompatibility of materials and unconventional details of this hybrid design.
Roof Drainage
The drainage system from the roof surfaces has funneled water unevenly, causing extreme erosion in certain locations, particularly around the canales and vertical wood drains.
Another design problems seems to have been related to insufficient sloping and drainage of the roof slabs. Much of the roof was too flat and caused ponding to occur.
The west wall has been unusually troublesome, particularly where the roofs of second floor rooms drain over the exterior walls. The east wall is freestanding and does not have adjacent second floor roofs. It is also more protected from prevailing weather and has therefore survived in the best condition.
The instability of the outer wythe of natural adobe is due in part to the failure of the roofing and drainage systems to protect the wall cavity from water intrusion.
Roofing Materials
The 1970’s choice of the roofing product, Bituthane has been a problem because of its fragility and extreme vulnerability to degradation by the sun. Bituthane was a new product on the market at the time and relatively untested.
Canales
Integral to the roofing system are the wooden canales which drain the roof surfaces. The poorly designed canales have not performed well and various modifications have been made over the years.
The canales were designed to project about 2 feet from the wall. Some were 10" to 12" diameter logs that were hollowed to create a gutter shape; other canales were assembled from three pieces of wood. The records are not clear on the extent to which the canales were lined with Bituthane. However, that roofing product would have cracked and failed almost immediately if exposed to the sun in the canales, or even at the edge of the roof at the canale. Several vertical wooden canales were installed recessed into the adobe walls. Proper maintenance of the canales would have consisted of removing soil build-up and treatment with linseed oil. In addition, the canales were inappropriately located in some cases, allowing water to drain on columns, beams, and the front gate.
Wall Construction
The system used for tying together the wall wythes has not been effective in producing a stable wall. Records indicate that the wythes of adobe were so unstable during construction that bracing was required. The inner wythe of Portland cement-stabilized block is tied to the outer wythe of natural adobe with modern sheet metal tie strips. This tie system failed due to rusting and lack of bonding with the natural adobe. A better tie design would have used adobe string courses and mud infill to create a monolithic wall, as in traditional adobe masonry wall construction. With such a monolithic wall, however, the difference in thermal expansion and contraction between the natural adobe and the stabilized adobe would cause some differential movement which could be problematic.
Additionally, if the fort had been designed in accordance with traditional, monolithic adobe wall construction principles, there would have been more solid mass to resist erosion.
Bond Beams
The reinforced lightweight concrete bond beams are continuous in all walls and over door and window lintels. All of the vigas, beams and lintels bearing on the walls are embedded and anchored with rebar in the concrete bond beam.
The design of the reinforced concrete bond beam does not allow for replacement of rotted vigas in the future.
Treatment of Wood Members
In many locations, the ends of the vigas extend beyond the bond beam through the external walls where they are exposed to weather. This exposed structural wood was not pressure-treated to resist decay, and no consideration was given to periodic replacement. Currently, the majority of the viga ends are in a state of major deterioration.
The wood members were stained a pale grey upon construction to give them a more aged and weathered appearance. According to the contractor's notes, the wood exposed to the exterior was treated with "Okon S2 and Tiffany" wood waterproofing agents; however, the specifications called for wood preservatives such as pentachlorophenol to have been used. All of these treatments were brush or spray applied. None of the major structural timbers was pressure treated. All of the interior wood members were treated with a flame retardant chemical, Flamort WC, according to the specifications. The contractor indicated that Thermo-Lag was the product used for fireproofing and Monokote 5 was sprayed in room 124, the mechanical room.
Use of Ferrous Metal with Adobe
Metal Conduit and Piping
The radiant heating system originally installed under traditional adobe mud flooring has been completely abandoned because the piping was corroded by contact with the adobe materials.
Similar problems have been caused by the use of metal electrical conduit in the damp, corrosive adobe walls.
Metal Wall Ties
The same problems of corrosion occurred with the metal ties used in the adobe wall construction.
Sheet Metal Flashing
Galvanized sheet metal flashing was installed where the original roofing abutted vertical surfaces. The Bituthane membrane was placed on top of the reinforced concrete roof deck. Over the Bituthane membrane was poured 3 inches of portland cement stabilized adobe mud, similar to the mix of the stabilized adobes, with a topping of pea gravel. According to Don Curry, former Chief of Maintenance, many sections of corroded sheet metal were removed during replacement of the roof. This is another instance of material incompatibility with the corrosive adobe mud and the sheet metal flashing in contact with each other.
MAINTENANCE PROGRAM
In the years since construction, this hybrid structure and its modern materials and systems have been vastly altered in attempts to correct design problems and maintain structural and historic integrity. Erosion of the adobe from intense spring and summer rainstorms has required re-rendering efforts far in excess of the maintenance program originally proposed. Massive maintenance efforts have been undertaken to halt a process of deterioration more extensive than anticipated by the 1976 designers.