Garden City Part I – Calculations

LittleCity1If it is the case that fossil fuels are scarce and not renewable on human time scales, then it follows that any sustainable or permanent culture must develop a food system that is independent from the use of fossil fuels. If one were to focus in on just agriculture, food production, in the context of no fossil fuels, what would it look like? It seems clear that production would be decentralized, intensive and predominantly local (probably more fresh, mineral rich and less wasteful, too). Is it too much to say that ideally the amount of food produced in any definable region would be at least equal to the amount of food consumed in that place (obviously, surplus would be needed for storage for resilience in times of poor production, and potentially for minor export from a place of optimal climate and soil)? In other words, might we need to produce as much food as we can as close to where we live as possible.

So, how much food should we grow in San Francisco? Aware of the fossil fuel situation above, many urban agriculture enthusiasts in the city say, “as much as we can?” Which begets the question, how much food can we grow in San Francisco? Let’s examine that question…

It seems like a simple problem to assess; catalogue the available resources, determine an average yield per acre or yield per square foot and do a little math. When one starts examining the process, one finds that any assessment becomes leaden with assumptions about resources, yields and human nutrition content of the yield.

One approach to assess “how much food can we grow” in any region would be determine a “ceiling” i.e., given a certain set of transparent assumptions, what would be the optimal annual production. Ideally then one would analyze yields by mineral, vitamin, fat, carbohydrate and protein content. Then one could begin to change the assumptions and discount off of the ceiling figure to begin to arrive towards a more “realistic” figure. Of course, I would advocate for starting with a ceiling approach because that would provide a more exciting target to aim for even if we never arrive at that destimation. Turns out even, arriving at a reasonable ceiling can be quite difficult due to the variables.

Ok, so how much can we grow in San Francisco. Any resident of San Francisco, likely knows that the limiting factor for food production in San Francisco is space – growable space. At 7 by 7 miles San Francisco has about as much land area in total as is currently vacant in the city of Detroit where urban agriculture seems to be permeating back into common culture. 30,000 acres of land, of which 70% of is said to be covered by impervious surfaces (rooftops, sidewalks, streets, parking lots, etc.). That leaves about 9,000 square feet not covered by a right of way or rooftop or other paved surfance. This raises a critical variable for determining our ceiling. Clearly some, many?, paved surfaces may be entirely appropriate for intensive food production in raised beds, etc. Likewise much of the 9,000 square feet of permeable surface might not be appropraite for food production. So what do we include or exclude in calculating our ceiling?

You may be immediately wondering about rooftops. In my assessment of San Francisco urban agriculture, I am going to mostly discounts rooftop agriculture. The reason I am excluding rooftops has to do with the great uncertainty about how much of our rooftop space is appropriate for intensive food production. This leaves us with a future scenario where we know we might be able to adjust upwards as we incorporate more rooftop gardening space.

OK, so back to the 9,000 unpaved acres. That land generally falls into two categories – publicly “owned” land and privately “owned” land and the curious anomoly of land controlled by the Presidio Trust. Publicly owned land includes all land owned by the city departments and agencies , including recreation and parks department land, public utilities commission land, department of public works (which includes right of ways – i.e., streets and sidewalks), redevelopment agency, mayor’s office of housing, fire department and more and state land controlled by CalTrans. Privately owned land includes commercial and residential property with front or backyards and privately owned vacant lots that are in the process of being developed or left vacant and unlikely to ever be developed. To assess how much of each type of land is available, we have researched a bit on using available data to arrive at the following table:

Land Classification (“Ownership”) Acreage (may include paved surfaces)
Recreation and Parks Department 3,466 (1)
Other Publically Owned “Open Space” 3,007 (2)
Presidio Trust 1,491 (1)
Backyards 5,500
Privately Owned Vacant Lots 108
CA Dept. of Parks and Recreation 252 (1)
GGNRA 175 (1)


As you may have noticed, simply classifying the land by ownership does not necessarily help us arrive at what portions of those acreages are “available” and “appropriate.” Both concepts, available and appropriate, are moving targets and at least somewhat subjective. Simple addition shows that there are 13,999 acres accounted for in the table above. Some of those acres are paved but potentially productive and some are unpaved and unlikly to be productive (slope, toxicity, etc.). Some may be available for interim periods between receiving development entitlements and actually development. Again, this calculation does not account for right-of-ways (streets and sidewalks) or rooftops. So what figure do we move forward with to arrive at a ceiling. Maybe if we turn attention to potential yield data we can then arrive at a few calculations about total possible production.

So, “how much food can be produced on one acre?“ (And, “how much food does it take to sustain a healthy individual?”). The amount of food that can be produced on any plot of land is variable based, at least, on available water, soil (or nutrients in the case of hydroponics), light and, oftentimes, slope or ability to intensively crop or maintain. Not a small set of variables and not easily averaged. Well, what about a ceiling in consideration of optimal conditions. This is also not a simple calculation (e.g., what is optimal?), but we do have some proxies for impressive yields from intensively managed gardens.

tomatoesFor example the Dervaes family of Path to Freedom in the favorable Pasadena climate, while using an abunadance of water, are able to grow over 7,000 pounds of food on 1/10th of an acre (4,356 feet). This does not account for how many calories there are in 7,000 pounds of food (Of course, it depends (energy density differs between fats, carbohydrates, etc.) on the type of food grown (e.g., Path to Freedom’s primary yields are vegetables and fruits with some protein from eggs and meat from small poultry), but it does arrive at an impressive number – 70,000 pounds per acre when intensively managed. Wow! That should provide a solid (really optimistic!) number for some ceiling calculations. Those calculations will not leave us with data that will help determine how many people we can feed in SF however. The question about how many calories per pound of food and how much food one person needs are troublesome.

A typical approach is to determine a daily caloric intake need and then determine how much land is needed to produce that amonut of calories. Of course, looking at caloric intake can be problematic. Different individuals of different height, weight, age and activity level need differening amounts of energy (calories) per day. Calories by themselves also do not account for full nutrition that begs us to assess mineral, vitamin, protien, carbohydrate and fat content of the food providing the calories. Some, such as John Jeavons, have summarized their research by reaching a figure of minimum land needed to support one individual (Jeavons puts it at .05 hectare or 5,382 square feet). Here are a few links where you can read more.

(3) (4) (5) (6)

Using a rounded Jeavons number of 5,000 square feet we can make some ceiling calculations on possible humans fed with full nutrition from food production within San Francisco. Of course, when we see these numbers I suspect we are going to desire an assessment of the entire SF Foodshed and peri-urban food production potentialities broken down by nutrition and crop energy density.

This might lead us to prioritize the production of certain types of crops in San Francisco – e.g., we may choose to optimize the production of highly perishable fruits and veggies within the city limits and focus on supporting soil-building, ecosystemic agriculture for grains, pulses, fats (from oils and animal protein) production in the peri-urban environment.

Ok, so how about some calculations to compare:



with discounted calculations (50% less ceiling acreage and 50% less optimal lbs per acre

Cultivated Acreage Yield @ 70,000 lbs per acre Yield @ 45,000 lbs per acre
13,999 979,930,000 lbs 629,955,000 lbs
9,000 630,000,000 lbs 405,000,000 lbs
4,500 315,000,000 lbs 202,500,000 lbs








with discounted calculations (50% less ceiling acreage and 50% less optimal sq. ft. needed per person

Cultivated Acreage
Humans Sustained @ 5,000 sq. ft. per
Humans Sustained @ 10,000 sq. ft. per
13,999 121,959 60,980
9,000 78,408 39,204
4,500 39,204 19,602







Obviously, we are not likely to feed San Francisco from within San Francisco at current population density, dietary habits, available space and other resources. In another post we’ll examine the process of assessing the viability of a foodshed so we can design how an evolving San Francisco fits into its bioregion. We’ve got some great references and data to look at, here are a few to get started:

Shelburne Falls Food Security Plan


San Francisco Foodshed


1 Response

  1. September 26, 2016

    […] Open land, obviously, comes at a premium in cities. And though an urban farm of one or two acres can produce a lot of food, the definition of “a lot” is relative, said Eli Zigas, food systems and urban agriculture program manager of the urban planning nonprofit SPUR. “It can be a substantial amount of food, but for a dense area, compared to the number of people who live nearby, it’s a small dent.” […]