Using Penciler to Analyze Sites for Affordable Housing Development
In the face of a widespread affordable housing crisis, housing developers and public housing agencies are under pressure to quickly assess vacant or under-utilized sites for multifamily affordable housing development.
City agencies located in California counties with affordable housing bonds must make decisions about where to allocate funding to maximize resources and provide high quality affordable housing. Like housing developers, city housing agencies need to perform a site feasibility analysis, or “test fit”, to determine the maximum number of units possible on a site given various constraints and requirements. Multifamily housing developers go through a similar process as they decide whether or not to pursue a new development project. Site feasibility analysis requires architectural, zoning and financial expertise and often requires a team of people to contribute, which makes it hard to iterate quickly and consider many different options. This process can be a bottleneck in the effort to quickly assess sites for new affordable housing development.
At UrbanSim, we participated in the San Francisco Startup in Residence program with the San Francisco Mayor’s Office of Housing and Community Development (MOHCD) to develop a web-based site feasibility analysis tool called Penciler.
Penciler was created not only to make the site feasibility analysis process faster and easier, but also to help users consider hypothetical zoning or policy scenarios and better understand possible outcomes. For example, an affordable housing developer might ask how many more units of income-restricted family housing would be possible on a site if they were to take advantage of the California State Density Bonus program. A city planner can analyze multiple sites in a corridor and understand the development feasibility impact of a zoning change. A market rate developer can consider the feasibility impacts of on-site inclusionary housing and play around with the building program to make the numbers pencil-out.
At UrbanSim we designed a proprietary layout algorithm and financial analysis tool to help users rapidly generate many different building options on the same site, visualize the options in context, and compare the cost and financing of each option.
This allows users to make a quick “go/no-go” decision about a site. Important zoning and building code information is stored in a database by city, allowing users to largely avoid interpreting a planning or building code. The algorithm works to automatically maximize the unit count given all of the constraints, testing a thousand building options in a matter of seconds.
We leveraged the expertise of staff at SF MOHCD, SF Department of City Planning and others to understand not only what is important in a site feasibility analysis, but also what requirements any new multifamily building must meet.
This includes requirements about a resident’s access to light, open space, stairwells and elevators, to name a few. Users can customize building program requirements to meet many different spatial and programmatic needs, from family housing to housing for the homeless, and everything in between. Users can consider a portfolio of potential sites over time or form an opinion on the initial feasibility of a single site in a matter of minutes.
Given ongoing policy debates in California around things like inclusionary requirements and up-zoning around transit, Penciler can also help housing policy-makers, advocacy groups and developers better understand and visualize the impacts of policy changes.
Over the coming weeks we will demonstrate different scenarios using Penciler to show how we can ask important questions about development feasibility for multifamily housing.
How do three different multifamily development options pencil out at 2051 3rd Street in San Francisco?
2051 3rd Street is a parcel zoned for Urban Mixed Use (UMU), which requires 40% of the units to be two-bedroom units. It has a maximum height of 68 feet with a 25% rear yard setback and an open space requirement of 80 GSF per unit. There is no specific density limit, although the density is restricted by the height, unit mix requirements and open space requirements.
We tested three different options on this site, all which meet the 40% two-bedroom requirement. We included 1,000 sf of ground floor retail in each option.
The first building option we created meets all zoning requirements; i.e. the height, setbacks, open space and the two-bedroom unit mix requirement. We also included 60% one-bedrooms in the building program. This produced a 55,200 sf, 65-unit, 6-story building in an L typology.
For the second building option we tested the assumption that we might get a zoning allowance on the 25% rear yard setback requirement. This allowed us to increase our unit count by 25 units for a 90-unit building. The following image shows the preliminary residential floor plan generated by the algorithm.
Lastly, because we are assuming that all of the units on-site are income-restricted, we experimented with the San Francisco local density program (SF-HOME). This allowed us to increase the height to 9 stories with 39 additional units of affordable housing.
In each building option we assume that the one-bedroom units are income-restricted at 80% of area median income and the two-bedroom units will be income restricted at both 60% and 80% of area median income. We also assume that each building will receive 4% Low Income Housing Tax Credits and an additional grant of $5 million.
The pro forma tool shows us that the building option using the local density bonus program has the lowest total development cost per unit, as the acquisition costs are spread across more units. It can also support the highest mortgage (because it has more income-generating units). The ‘Gap’ shows us how much subsidy is required to make each building option pencil (a necessity in San Francisco due to high construction costs and restricted rents in affordable buildings).
We can look at these three options from different perspectives and consider different priorities. There are many factors to consider, but if the priority is to maintain the height and setbacks required by the default zoning, we may need to play around with the income levels of the units to achieve a more financially sustainable building in the long-term — the yield on cost for this option is around 2.9%. If the priority is to maximize the unit count on the site and decrease the subsidy needed per unit (although not the total subsidy), we might look to the density bonus option.
Over the coming weeks we will show additional examples of multifamily development feasibility analysis using Penciler. If you like this story, please clap and share with others who are working to address the affordable housing crisis. Penciler is currently in beta testing in the Bay Area. To find out more about Penciler or to request a demo, visit www.penciler.org or contact us at info@urbansim.com.