Department of Architecture

Center for Housing Innovation

Donald Corner, Director

Experimental Student Family Housing
Technology: Design, Production and Resource Conservation
Accepted for presentation at the ACSA Annual Meeting, Seattle, WA, March 1995.

Donald B. Corner, Associate Professor
William G. Sturges, Adjunct Assistant Professor
Center for Housing Innovation, University of Oregon

One Story Building | One & a Half Story Building | Two Story Building

Synopsis:

This paper reports findings from the first of two phases in a multi-year research project. The project includes design and construction of new housing prototypes followed by long term energy performance monitoring of the completed units. The prototypes e mploy various means of industrialized production and the overall project explores the role of these techniques in the realization of quality, affordable housing units which conserve both material and energy. The project also serves to showcase regional e fforts to add value to a declining timber resource through the production of integrated building components.

Systematic design investigation yielded a full spectrum of prototype possibilities which were subjected to both qualitative and quantitative evaluation. The most promising unit types were prepared for competitive bidding among manufacturing firms which em ploy a range of prefabrication methods. Bids from site builders were also collected as a basis for cost comparison. For each of the unit types the most effective combination of techniques was selected for construction.

The completed units successfully demonstrate the application of new production methods to high performance, energy conserving dwellings. Substantial savings in construction cost and annual operating expense were realized through these process innovations . However, the project also demonstrates the unrealized potential for savings through reform of land use standards and the development process. The ultimate goal of the project was to promote local industries and prepare them for entry into highly compe titive export markets, and this was a resounding success.

Objectives of the Project:

Our university is located in a region historically dependent on timber production as a base of employment and economic activity. The research described in this submission is part of an effort at state and local levels to stimulate new businesses capable o f adding value to a declining timber resource while creating new family wage jobs in manufacturing. One of our specific roles in this effort has been the design and construction of six units of family housing intended to showcase the capabilities of new i ndustrialized housing producers in our region. The project also serves a national agenda. The research was supported in part by federal funds committed to the development of energy efficient housing built with new materials and industrial technologies. The initial project goals were as follows:

Ą

Test the role of industrialized building systems in the production of quality, affordable housing.
Ą

Explore the relationship between these industrial production methods and emerging standards of energy

efficiency for the completed units.
Ą

Test the capability of manufacturers in the region to generate cost effective building components and

systems.
Ą

Promote successful prototypes through documentation of the construction processes, analysis of labor and

material costs and measurement of the energy performance.

Construction of the model units was financed through state bonding authority granted to the university for development of a much larger number of student family housing units elsewhere in town. The six experimental units were assigned to an area immediate ly adjacent to the campus in which university owned residences stand on 60-80% of the lots. Given the user group and real constraints of the site, the project was expanded to include additional issues:

Ą

Test local planning instruments as the means to increase residential density within an older, single-family

neighborhood.
Ą

Demonstrate simple concepts of site design: private outdoor spaces, common play areas and solar access.
Ą

Explore unit configurations which are efficient, livable and adaptable to special needs.
The program statement called for five two bedroom units with a sixth to include an extra bedroom and bath. This became an opportunity to explore expansion of a small unit within the constraints imposed by the industrialized building techniques.

Research Design and Methods Used in the Research:

The prototype design process was organized in terms of a matrix. One axis ordered the plan types by the level of industrialization, increasing from simple components (trusses) to open panels, closed panels, core units and full modules. The other axis beg an with single, double and triple unit buildings. As site design alternatives were explored, it became clear that double units would generate the best fit. Similar units in pairs also offered greater control of the energy use monitoring program planned f or later phases of the project. Therefore, the matrix was expanded to examine 1 story, 1-1/2 story and 2 story prototypes within the double unit category. Schematic unit prototypes were developed for each cell of this matrix with a bias toward the higher levels of industrialization (modules) to differentiate the proposals from conventional practice. Once developed, the prototypes were sent out for review by relevant manufacturers (panel, module, etc.). Through this cooperative dialogue we began to gain a sense of what was technically possible and cost effective. With support from an allied research team, we performed a schematic energy analysis of each prototype. As a consequence of these evaluations we selected a range of solutions which combine site- built thermal mass with off-site manufactured elements. Light weight panel and modular units are already widely available, and the addition of thermal mass is a very effective energy strategy in our region. Large areas of south glazing capture some solar savings in our relatively mild winters, while interior daylighting is significantly improved over that which energy optimization would permit without added mass. On summer evenings, prevailing north winds will cross-ventilate the units, cooling the mass for the following hot afternoon.

We began the long process of site design approval concurrently with the prototype studies. After careful examination of all the local planning ordinances we elected to develop the units as single family townhouses within a cluster subdivision. This prove d to be the only means of increasing the density without a politically volatile zone change. Lot lines internal to the cluster require no setbacks, allowing the units to be attached. Solar access for both the new and existing units suggested placement of the 2 story, 1-1/2 and 1 story units in a cascade descending toward the north, so that the longest shadows fell within our own building group.

In order to boost density on the demonstration site (eight units, with two existing, on three lots), we imbedded it in a larger cluster of 15 units which meets the area requirement of the present zoning. As a by-product we were able to create lots for old er houses relocated from other university construction sites. These fill in the missing ŇteethÓ in the residential block.

Working continuously with industry, we developed generic prototypes for the unit pairs that could be built with either panelized or modular methods. Following two rounds of professional cost estimating, bid documents were prepared reflecting what appeared to be the most effective techniques:

Ą

1 story: closed panels with a preference for structural insulated panels for both the walls and roof.
Ą

1-1/2 story: open panels combined with attic trusses for the upper floor.
Ą

2 story: modular second floor over a thermally massive first floor.

Data Collection and Analysis:

A local contracting firm was selected to supervise the bidding and building process within a construction management/general contractor format. Public bids were collected on all the building components and subsystems with alternates organized in a fashion which permitted direct competition between different manufacturing techniques. After a thorough review of the bid data we put together a revised production strategy:

Ą

The 1-1/2 story unit was awarded as anticipated to a local firm specializing in open panels for both exterior

and interior walls. The firm make extensive use of CAD and has a relatively low-tech production facility which

easily accommodated the dormer s and gables in this structure.
Ą

Only insulated slabs on grade and concrete masonry party walls proved to be cost effective as thermal mass

in our climate. Thermal mass on exterior walls was dropped.
Ą

The 2 story units were awarded as wood stud closed panels from a local firm using a highly automated factory.

The modular bids were not competitive for a number of reasons.
Ą

The 2 story units also utilized prefabricated floor cassettes and roof trusses, permitting us to explore the

structural, plumbing and electrical integration of these components with the wall panels.
Ą

Foam core structural panels proved to be significantly more expensive in this application, particularly as a roof

assembly. Therefore, the 1 story unit pair was set up as a side-by-side comparison between foam core panels

on one half and wood stud clo sed panels on the other. Identical truss roofs cover both units.

The individual unit designs and the respective construction methods are described and illustrated in the supplementary materials.

Construction of the project proceeded following the acceptance of a guaranteed maximum price (GMP) by the university. Throughout the process detailed time and materials records were kept by the CM/GC. The design/research team maintained a daily log of pr oblems and future opportunities as they were uncovered in the field. Extensive photo and video records were taken at each step in the sequence. We are at present working with the CM/GC to complete a retrospective cost analysis to accurately assign the co st of field changes to those sub-systems which caused the problems. For example, the final cost of closed panel systems must accurately reflect problems encountered by the plumber and electrician. However, such changes will be relatively small, and it is already possible to draw conclusions from the recorded costs of the work.

A second phase of this research project began as soon as construction was completed. A series of energy performance tests have been conducted including blower door tests, infrared camera scans, tracer gas tests and simulated occupancy. The units include low voltage sensors and electrical submeters which permit long term monitoring of the ambient climate variables, internal and external surface temperatures and energy use of space heat, hot water and electrical devices. Occupied by student families, the u nits will be monitored for a period of two years. At the conclusion we will be able to finally compare the actual energy performance with computer modeling completed during the design process.

Major Findings and Results:

The conclusions summarized in this section refer back to the project objectives. However, to simplify the narrative the two lists above have been combined and reordered.

Residential density on the demonstration site has been increased while maintaining pleasant, sunny private spaced both inside and outside of each dwelling. The average land area assigned to each unit, inclusive of parking and common space, has been reduce d to 70% of that required by the present R1 zoning and only 37% of the pre-existing pattern. Critical to this success was a new city ordinance which permits one parking stall per unit for student housing. Our qualitative design criteria and setbacks for solar access did not permit us to reach our initial goal of R2 lots at 50% of the R1 area. Nevertheless, the central play area and shared laundry facility should support a strong sense of community which was a requirement of our client.

Spaces within the units are efficient and useful. Elongation of the unit plans east to west for solar gain results in longer indoor views across the main living spaces, and the additional glazing area permitted by the use of thermal mass provides generous and welcome daylight in most rooms. The one story unit easily accepted the constraints of barrier free design, although the larger corridor widths do reduce efficiency. Visitors to the project have informally expressed an overwhelming preference for the 1-1/2 story units with their traditional appearance. The two story units are strikingly tall with extra roof thickness for insulation, although in terms of land area and construction material, these are clearly the most efficient.

Industrialized building techniques did make a meaningful contribution to the quality and affordability of the finished units. The most experienced producer completed the 1-1/2 story frame and enclosure for $2500 less than the baseline, site-built bids. T he 1-1/2 story shell was completed in only three days, while the 2 story exterior reached the second floor top plate in only one day. Given the public financing of this project, no direct savings resulted from the reduced project schedule. However, such savings would accrue in private developments. The most significant issue in public work is the wage differential between the site and factory under the Davis-Bacon Act, contributing more to the factory advantage than frequently overstated productivity gai ns.

The limiting factors in making truly affordable infill units have been development costs related to the land and not the unit production. The planning ordinances in our town are still written around the conversion of raw land to low density subdivisions. New instruments are needed to encourage small scale infill within existing neighborhoods. At present the entire regulatory system presumes the existence of large, profitable developers who can and must be compelled to install appropriate public facilitie s. Single houses on single lots escape such scrutiny. A small, multiple unit project like this one carries a disproportionate burden of direct costs and added staff time required to coax it through a long approval process. The cluster subdivision opene d the door for a seemingly endless list of requirements for the infrastructure at costs which far outweighed the savings realized through industrialized building.

The energy performance goals of this project appear to have been met, although more will be known at the end of phase two. A range of manufacturing techniques were used, and all of the units meet the highest levels of the Energy Good Cents incentive progr am sponsored by our regional power authority. This places them well above the most strict current codes. Completed tests reveal the units to be tight and well insulated, and the thermal mass was integrated without significant detailing problems or cost p enalties. The specific contributions of industrialization are subtle: windows installed in the factory fit well; insulation is more accurate, and the layer of rigid foam board under the sheathing costs much less to install in the factory. Manufactured co mponents like foam panels without studs and high heel trusses make significant contributions, although these are also available to site builders.

New manufacturers in our area learned a great deal from their participation in the project. The contrasts in their techniques became clear as we progressed. The open panel producers use a very low tech approach in the factory with ingenuity, construction experience and efficient computer usage replacing high capital investments in equipment. As a result, they have great range in what they can panelize and demonstrated savings on dormers, interior partitions and stair cases, in addition to the basic shell . Keeping overheads low and leaving plumbing, wiring and sheetrock to traditional methods positions them to compete aggressively in our region where on-site costs are still somewhat reasonable. The closed panel builder has a deep investment in equipment and depends on a high volume of highly finished components. The equipment limits design flexibility, and time taken to complete special elements off the line does not help to amortize loans on the plant. With our unit prototypes they were able to reach c ost levels which give them easy access to markets in which site based finish work is much more expensive and energy efficient design is of similar priority.

The head to head comparisons of the 1 story units were of particular interest. The closed panel half went together very quickly and for the present at a much lower material cost. Siding, sheetrock, windows and trim went on the foam core panels in the fie ld, and the electrician reported a threefold increase in his labor required to complete the rough wiring. The experience of our energy analysis group indicates that long term savings will accrue from stable insulation without thermal bridges, but it will be difficult to offset the negatives. Nevertheless, the general contractorŐs forces based on the site had a strong preference for working with the foam core panels, because they felt more involved in traditional hand assembly.

Significance and Application of the Results:

The lessons we have learned from this experiment have already been presented at a national conference on urban design, the NAHB Building Systems Council and to individual manufacturers around the country. Further papers have been accepted detailing the en ergy performance and monitoring aspects. Our research team works cooperatively with state economic development personnel and the housing division on topics of mutual interest. We have made the case to the governorŐs office that truly affordable housing a waits a hard look at the priority given to land development costs at the expense of the building stock. This is, of course, a national issue.

The principal investigator on this project has a new grant of support from the state to provide continuing consulting to emerging industrialized builders and component manufacturers. There remains a great deal to be done, particularly in the area of testi ng and approval of closed panel systems for ready acceptance throughout our state and in others. Manufacturers also need assistance with the development of prototypes for more demanding climate zones and for multi-family densities.

Nevertheless, the ultimate goal of the project, to develop and promote local industry, has been a resounding success. Critical construction details were developed through the cooperative efforts of the research team and the producers. The manufacturing e xperience and cost confidence acquired along with visual documentation of the assembly process and site tours of the completed units have thus far resulted in a major contract for export of 200 units per month to Japan -- no longer raw logs, but closed wal l panels and floor cassettes comprised entirely of American materials and labor.

Arch. Dept. Research | Architectural Computing | DIL Foyer | The DIL References | DIL Index

Web posting 96.06.15-KMM