A Novel Approach for Calculating the Feasibility of Urban

Agriculture using an Enhanced Hydroponic System

Item Type text; Electronic Thesis

Authors Kacheris, William

Publisher The University of Arizona.

Rights Copyright © is held by the author. Digital access to this material

is made possible by the University Libraries, University of Arizona.
Further transmission, reproduction or presentation (such as
public display or performance) of protected items is prohibited
except with permission of the author.

Download date 01/07/2020 05:08:04

Link to Item http://hdl.handle.net/10150/620674

 A NOVEL APPROACH FOR CALCULATING THE
FEASIBILITY OF URBAN AGRICULTURE USING AN
ENHANCED HYDROPONIC SYSTEM

by

William Kacheris

A Thesis Submitted to the Faculty of the

DEPARTMENT OF AGRICULTURE AND BIOSYSTEMS

ENGINEERING

In Partial Fulfillment of the Requirements

For the Degree of

MASTER OF SCIENCE

In the Graduate College

THE UNIVERSITY OF ARIZONA

2016

1
 STATEMENT BY AUTHOR

The thesis titled “A Novel Approach for Calculating the Feasibility of Urban
Agriculture using an Enhanced Hydroponic System” prepared by William Kacheris has
been submitted in partial fulfillment of requirements for a master’s degree at the
University of Arizona and is deposited in the University Library to be made available to
borrowers under rules of the Library.

Brief quotations from this thesis are allowable without special permission, provided that
an accurate acknowledgement of the source is made. Requests for permission for
extended quotation from or reproduction of this manuscript in whole or in part may be
granted by the head of the major department or the Dean of the Graduate College when in
his or her judgment the proposed use of the material is in the interests of scholarship. In
all other instances, however, permission must be obtained from the author.

SIGNED: William T. KacherisWillia

This thesis has been approved on the date shown below:

Gene Giacomelli Date

Thesis Director

2
 ACKNOWLEDGEMENTS

I would like to thank Dr. Giacomelli for his advice and help in the creation and execution

of this research and for giving me the opportunity to learn and grow at the University of

Arizona for the past two years. I would also like to thank the committee members Dr.

Yitayew and Dr. Tronstad for all the time spent helping me to form a rough idea into a

thesis and mentoring me towards a final, realistic goal.

I must also acknowledge Dava Jondall for all of the assistance throughout my two years

at the University of Arizona in both advising and research. Charlie DeFer and Don

Clifford, and Juan Gonzaleaz also deserve thanks for all of the time spent in the shop

constructing the a-frame. Thank you Neal Barto for your patience and countless hours of

help in the greenhouses.

Special thanks to Ken Hickman, Brooke Conrardy, Ashley Hanno, Caitlyn Hall, and

Isaac Hung for helping with the research in multiple ways.

3
 DEDICATION

To my brother Matt, my mother Maribeth, and my father Peter for their incredible

support during the creation of this thesis, none of this would have been possible without

you.

4
TABLE OF CONTENTS

List of Tables……………………………………………………………………….........06
List of Figures……………………………………………………………………………07
Abstract..............................................................................................................................10
Introduction………………………………………………………………………………11
Problem Statement……………………………………………………………………….12
Literature Review………………………………………………………………………...13
Modern agriculture……………………………………………………………….13
Introduction, description of urban agriculture…………...…………...….13
The state of urban ……………………………………………………………..…14
Local food………………………………………………………………..14
Success for local food in urban markets…………………………………15
Shanghai success story……………………………………………….…..16
Cuba urban agriculture Success…………………………………….……16
Barriers to Entry……………………………………………………………….…17
Introduction……………………………………………………………....17
Zoning Restrictions. ……………………………………………………..18
Distributing of urban produce…………………………………………....18
Regional Food Hubs...…………………………………….....…………..19
Tax Incentives…………………………………………………………....20
System Designations………………………………………………………...…...21
Introduction…………………………………………...……….…………21
Definition of hydroponics………………………………………………..22
Choosing a correct hydroponic system…………………………………..22
Growing area Concerns………………………………………………….25
Greenhouse Considerations……………………………………………...25
Objectives………………………………………………………………………..26
Materials and Methods………………………………….…………………………….....27
Hydroponic a-frame design/construction………………………………...............27
Specifications.............................................................................................28
Crop Trials.............................................................................................................32
Greenhouse................................................................................................33
Transplant Procedure…………………………………………………….33
Data recording……………………………………………………………37
Proving validity of the system …………………………………………..38
Economic Model……………………………………………………………..39
Auxiliary Sheets....……………………………………………………….40
Heating..............………………………………………………….40
Labor sheet.....................................................................................41
Nutrient solution sheet...................................................................42
A-frame build cost sheet................................................................42
Cooling/Electrical sheet.................................................................42
Additional materials sheet..............................................................43
Main Page………………………………………………..........................43
Inputs..........................................................................................................44

5
 System Assumptions..................................................................................46
Facility Designations………………………………………………….....46
Crop output………………………………………………………………47
Cost Breakdown………………………………………………………….47
Investor tools..............................................................................................48
Results/Discussion…………………………………………………………….…………50
A-frame..................................................................................................................50
A-frame capabilities.………….......…………………………….….…….50
System improvements………………………………………….………...51
Crop trials...............................................................................................................53
First harvest data.………………………………………….……………..53
Second harvest data....................................................................................55
Third harvest data......................................................................................59
Fourth harvest data.....................................................................................62
Modeling…………………………………………………………………………66
Model developed for use by potential investors…………………………66
Clarity from auxiliary sheets..........................……………………………67
Heating cost sheet......…………………………………............................68
Cycles as growth-time measurement.........................................................68
Investor tools..............................................................................................68
Year 1 analysis...............................................................................69
Example inputs….....……………………………………………..69
Sensitivity analysis…………………………………………….....69
Net cash position............................................................................69
Conclusion……………………………………………………………………………….74
Objectives………………………………………………………………………..74
System Design and Experiments.............………………………………………..74
References............................………………………………………………………….….75
Appendix............................................................................................................................81
Jensen Formula........…………………………………………………………......81
A-frame construction build list..............................................................................82
Pump sizing calculations........................................................................................83

LIST OF TABLES
Table 1.........……………………………………………………………………......……54

Table 2.........……………………………………………………………………......……54

Table 3.........……………………………………………………………………......……56

Table 4.........……………………………………………………………………......……57

Table 5.........……………………………………………………………………......……61

6
Table 6.........……………………………………………………………………......……62

Table 7.........……………………………………………………………………......……64

Table 8.........……………………………………………………………………......……64

LIST OF FIGURES
Figure 1: A 3-D computer rendering of the a-frame design using Solid Works rendering
software………………………………………………………………………......28

Figure 2: A view of the East face of the a-frame hydroponic system. Note the holes along
the length of each pipe and the first-generation reservoir, later upgraded to a
larger tank…………………………………………………………………..........29

Figure 3: A view of the South face of the a-frame system still under construction,
highlighting the aluminum struts forming the support structure of the system
which all components are attached to……………………………………………29

Figure 4: The Northern face of the system, the reservoir of the system is placed within the
empty space created by the a-frame design. The 416 liter reservoir is insulated
with a solar wrap and Styrofoam blocks to limit solar heat gain and to prevent
algae growth within the tank…………………………………………..................30

Figure 5: View of the return line bulkhead into the reservoir from the North face, a
13 mm diameter union was placed in-line to allow for simple servicing..………32

Figure 6: An example of a ‘Rex’ Lettuce seedling grown in Jiffy Preforma material, nine
day growth from start of seeding. ……………………………………………….35

Figure 7: The Jiffy Preforma growing material trays were germinated and grown to an
age of 14 days inside the environmental growth chamber at the University of
Arizona Controlled Environment Agriculture Center............................................35

Figure 8: The seedlings were placed at a height of .45 meters below the light fixture after
the initial twenty four hour period inside the growth chamber..............................36

Figure 9: An example layout of the input section to the main sheet..................................44

Figure 10: The assumptions of the system as they appear on the main sheet....................46

Figure 11: Facility designations of the model which are calculated outputs of previous
inputs, on the main sheet. ………………………………………………..............46

Figure 12: The display of the crop section after calculating a given set of inputs............47

7
Figure 13: A typical output of the variable and fixed costs portion of the cost
breakdown on the main sheet. ……………………………………………….......48

Figure 14: An example sensitivity analysis, note the yield percentages along the y axis,
market prices along the x axis, and the profit-per-cycle as the response...............49

Figure 15: A view of the Eastern face of the system, the inline filter can be seen along
the main riser, above the pressure bypass valve and pump. One may also note the
pressure-regulating distributors along the diagonal aluminum frame of the system
with individual feeder tubes directed into individual channels.............................50

Figure 16: The absorption of macro and micro nutrients varies with pH of the solution,
particularly, several micro nutrients such as iron and manganese become nearly
‘locked out’ at a pH above 7.0. (Kujawski, 2014).................................................52

Figure 17: Graphical representation of the first harvest fresh weights. Each bar
represents one of the ten samples taken from each row, each grouping of bars
represents the weights of a whole row. Ten samples were used in the analysis per
row. …………………………………………………………………………...…55

Figure 18: Graphical representation of the first harvest fresh weights. Each bar
represents one of the ten samples taken from each row, each grouping of bars
represents the weights of a whole row. ………………………………………….55

Figure 19: All heads under the designation of ‘crop #’ that are labeled with a ‘2’ were
from the second experiment and their placement is designated in the above
excerpt from the crop layout document. The spacing style (Single or Every Other)
and the subsequently required number of heads (10 or 19) is also
specified.................................................................................................................58

Figure 20: Fresh lettuce head weight for 10 samples (1 – 10) from the second experiment
for each row number (2 through 10), and each face (East (E) or West (W)).The
colors within each grouping represents the different rows....................................58

Figure 21: Dry lettuce head weight for 10 samples (1 – 10) from the second experiment
for each row number (2 through 10), and each face (East (E) or West (W)). The
colors within each group represents the different rows.………..........………......59

Figure 22: Fresh lettuce head weight for 10 samples (1 – 10) from the third experiment
for each row number (7 through 16), and each face (East (E) or West (W)).The
colors within each grouping represents the different rows....................................60

Figure 23: Dry lettuce head weight for 10 samples (1 – 10) from the third experiment for
each row number (7 through 16), and each face (East (E) or West (W)). The
colors within each group represents the different rows.........................................61

8
Figure 24: Fresh lettuce head weight for 10 samples (1 – 10) from the fourth experiment
for each row number (2 through 10), and each face (East (E) or West (W)). The
colors within each grouping represents the different rows....................................65

Figure 25: The ANOVA output of the SAS software comparing equal means among
spacing arrangements, fresh weight.......................................................................65

Figure 26: The ANOVA output of the SAS software comparing equal means among
spacing arrangements, dry weight..........................................................................66

Figure 27: The input area of the main sheet of the feasibility model. Sixteen input points
allow the user to form an accurate model of a hypothetical urban agriculture
operation................................................................................................................70

Figure 28: The facility designations area outlines the basic functions calculated from the
input area which feed later into the model. The crop section, both in cycle and per
year, give the user a clear readout of capabilities of the hypothetical facility.......71

Figure 29: Cost analysis calculations of the feasibility model under a hypothetical
situation models both variable and fixed costs of an operation.............................71

Figure 30: The revenues, profit and year 1 analysis portions of the feasibility model give
the user an area to adjust the market price of a head of lettuce to find an
appropriate profit margin. The year 1 analysis gives a breakdown of profit and
cost areas within the operation on a per unit basis.................................................72

Figure 31: The sensitivity analysis of the feasibility model for a hypothetical urban
agriculture operation. Allows better perspective to the user for determining
growing space and price point as the two major influencers of profit...................72

9
ABSTRACT
With a continued worldwide trend in population shift from rural to urban areas predicted

to increase, new approaches to agricultural production must be considered and

implemented. Little academic interest has been applied to determining economically

viable urban agriculture crop production sites for business investment. A feasibility

model to aid investors in selecting appropriate sites for the development of urban

agriculture food production within population centers was created. Lettuce crop trials

were performed from August 2015 to December 2015 at the University of Arizona

Controlled Environment Agriculture Center to validate the productivity of a unique high

density hydroponic system designed for the rooftop environment. The feasibility model is

based on this system and with a minimal number of inputs, ranging from size of growing

space to growing media costs, determines a wide range of useful outputs. These outputs

include crop productivity within the facility, material inputs and a cost breakdown of

starting a new agricultural venture. The model utilizes multiple sheets within one excel

document to give the user a clear and organized financial perspective of a hypothetical

growing operation in the main sheet. With this model, investors into urban agriculture

will have a means to gain an objective view of financial considerations before substantial

investment is completed.

10
INTRODUCTION

Sustainable food production and delivery has become a necessity of the modern

world. The traditional agricultural structure, involving massive logistical operations to

deliver produce into large population centers, uses enormous energy inputs and is

becoming less efficient for a world with half of all citizens living in cities (Mengual,

Esther, et al., 2013). Almost all fresh produce consumed in city centers is produced in

distant agricultural regions and transported via overland trucking or shipped across

oceans (Getter, 2006).

As people have become more aware of the source of their food, consumer trends

have shifted, allowing for locally grown produce to be desired by a sizeable portion of the

population. This change in buying pattern has been addressed through urban agriculture

operations of varying sizes and scopes, the aim of which is to deliver fresh local produce

directly to customers, all with minimal transportation costs. According to a report done

by the USDA on trends in local and regional food systems, fresh produce buyers are

interested in locally grown foods. Not only the source of the food interests consumers but

also how it is sold to them. Between 2006 and 2014 the number of farmer’s markets in

the United States increased 180% to a total of 8,268 while the number of direct to

consumer (DTC) sales increased 32% from 2002 to just 2007. Direct to consumer sales

are defined, in terms of fresh produce, as any transaction occurring between the producer

and the consumer which eliminates the distribution level, such as farmer’s markets or

road-side stands. More growers are joining the industry to reap the benefits of a sector

which the USDA reported the total value in 2012 of $6.1 billion (Low, 2015).

11
 Market interest from investors and governments has increased because of new

consumer trends. However, little research has been done to address the specific costs and

feasibility associated with an urban agriculture operation and its difference from

traditional agricultural models (Mengual, et al., 2013). An economic model which is

intuitive enough for an investor, who may not be from the agricultural field, to use while

also allowing an experienced grower options to adapt to different locations will be

helpful for the emerging urban agriculture sector.

12
 LITERATURE REVIEW

Urban Agriculture (UA)

As the population of the world grew from one billion persons in 1804 to over

seven billion people today, the food supply has been struggling to match pace (Leisinger

et al., 2002). To meet the demand of the population, farmers shifted into large scale

agricultural techniques, employing thousands of acres and the newest technology to

increase efficiency and yield. The crop production efficiency rose exponentially as

modern and industrial farming techniques overtook traditional, subsistence farming

practices. However, to generate food for the rising population, farms could no longer

operate within population centers and the rural/urban separation we see today, with

agricultural production sites often hundreds of miles from large cities, began to take

shape. The distance from field to consumer created the need for new and massive

distribution networks. Much more efficient than previous farming techniques, the new

production methods did incur logistical costs that are realized both in monetary and

environmental terms (Mengual et al., 2013). Although limited in the developed world,

this titanic distribution network results in the most urban, often times poorest, inner city

areas not receiving proper access to fresh produce at reasonable prices, these are referred

to as ‘food deserts’. Urban agriculture is one possible solution to this problem that can

bring the fresh produce back to the city center and reduce logistical costs caused by long

distance transportation.

Hendrickson defines urban agriculture as “The growing, processing, and

distribution of food and other products through intensive plant cultivation and animal

husbandry in and around cities” (Hendrickson, 2012). It can involve the use of soil or

13
hydroponic media for plant production, and urban agriculture encompasses all plant and

animal production, as long as it is within the city. Urban agriculture is most often

distributed locally, with the clear end point being either DTC sales or marketing to a

larger local consumer base through food hubs or full-scale distributors who are catering

to local businesses/restaurants (Low, 2015).

Many operations that serve the city markets practice organic growing techniques

to entice a consumer base which is willing to pay more for these specialty items due to

the perceived health benefits with organic certification. While most practice sustainable

growing, an organic certification is unobtainable for a new company unless there is

significant investor backing. According to a 2011 study by the Food Marketing Institute,

knowing the source and production style of an agricultural product was the largest factor

for 40% of consumers in their decision to buy locally grown food, second only to

freshness of local food (Food Marketing Institute, 2011).

The State of Urban Agriculture

Popularity for local food has been growing steadily over the past two decades.

Research into the field has been relatively slow but larger government interest is altering

that trend (Low, 2015). Research has been primarily devoted to case studies involving

small urban farms in countries outside the United States, and usually in developing parts

of the world. These foreign farms are useful to researchers to demonstrate how a food

system can benefit from the addition of local agricultural production sites near population

centers. Another portion of studies focus on locally grown foods that are delivered to the

14
consumer through DTC sales, farmer’s markets, food hubs, or even grocery stores in the

United States. The term ‘locally grown’ can encompass urban agriculture while also

representing large farms that sell direct to their consumers and small traditional farms

that operate outside of major distribution networks. This section will examine current

trends in the urban agriculture marketplace, how those trends came about, and how

efficiently farmers are addressing these trends.

Consumers are interested about the source and production method of their fresh

produce. The USDA reports that both the value of local food sales and the number of

DTC sales are increasing. With an estimated value of 6.1 billion in 2012, the local food

industry has more consumers and more producers than ever before (Low, 2015). It is

accepted that local food supply chains have higher per unit costs than traditional

production means. However, these operations can function efficiently through close-

market logistics and eliminating transportation and spoilage costs. Marketing based on

the production style (i.e. organic, pesticide-free) in addition to the crop being grown

nearby the consumer, can help growers to overcome higher production costs by

increasing the cost of their product. Crop diversification, such as having a variety of

atypical products, is another strategy employed by successful local farmers to reach a

larger market share (King et al., 2010).

Local food demand is highest among persons in urban areas, with the number of

local food producers continuing to increase in cities. Due to a rising number of DTC sales

opportunities and improved infrastructure for local producers in high population areas,

the demand for local produce is beginning to be met (Lichter and Brown, 2011). There

has been research done on a number of urban agriculture locations across the globe, all

15
with different motivations and solutions specific to their area in regards to crop

production.

Shanghai grew during the 20th century into a city of 24 million citizens,

agriculture was pushed away from the city center as land prices increased and

development pressed outward. To combat the urban sprawl, the local government began

an initiative in the 1990’s to increase agricultural production within and surrounding the

city. The local government utilized high capital investment in new and efficient

agricultural practices to encourage city-wide participation. The city also proposed the

idea as a means to have cleaner air and more green spaces within and near the city (Yi-

Zhang et. al, 2000). In the following years, UA production has increased heavily with a

majority of the items coming from the outer circumference of the city limits while still

containing a thriving community-level garden program in the inner city. With demand

high and food security risks abundant, mega cities such as Shanghai are heavily involved

in urban agriculture, with the United Nations declaring the city ‘self-sufficient’ as it

produces almost all of its agricultural products within its limits (Moreno et al., 2008).

Cuba is a small nation in land mass, and with limited imports, it must utilize

urban agriculture as a necessity. With an embargo maintained by the United States for

over fifty years and the fall of its ally the Soviet Union in 1989, Cuba embraced a

sustainable approach to agriculture, with government backing for projects that allowed

citizens to grow their own produce and sell it using communal gardens and programs.

One example of these programs is the ‘organoponicos’ system of government built,

raised growing beds that are rented out by growers and are located near city centers and

in areas of poor soil. The organoponicos system is just one of many different

16
government-supported strategies used by the Cubans for urban agricultural production

(Taboulchanas, 2001). In a study on the effectiveness of urban agriculture in Cuba,

researchers found that a society can embrace urban agriculture and produce up to 30% of

their food supply through backyard plots, community gardens and other forms of UA

production and to overcome food, material, and land shortages (Sanchez, 1997). Urban

agriculture has emerged as a viable option for more secure food sources even with

cultures as different as Cuba and China.

In recent years, participation in urban agriculture has increased worldwide, with

the developed countries of the world expanding rapidly while under-developed portions

of the world have remained fairly steady in their agricultural practices near and within

city limits. There are approximately 200 million people currently growing agricultural

products within urban areas, accounting for nearly 20% of the total global food supply

(Armar-Klemesu, 2000). The United Nations estimated that over 800 million people are

actively engaged in urban agriculture, either producing, buying, or distributing, and they

predict this number to rise as urban areas become more and more populated (Mendes et

al., 2008).

Barriers to Entry

New constraints are placed on the UA grower that restrict participation. Large

cities have much stricter zoning laws than in rural areas, and in addition property values

are magnitudes higher. Production materials are also much harder to supply to an urban

based operation due to the natural congestion of the city zone and therefore the profit

margin of the final product becomes less.

17
 A major obstacle to urban agriculture has been urban planning and zoning by city

governments. Although such laws are useful for the protection and the realities of

business operations within cities, they can be hampering to new industries. A study was

undertaken in the sub-Saharan city of Dar Es Salaam, Tanzania to determine how much

effect the expanding population and land area of the town had on food security. As Dar

Es Salaam grew, agricultural production centers began appearing throughout the

unregulated zones of town and a new food supply network emerged. From the addition of

the new farms, a more stable food supply developed and new zoning protection from city

officials aided farmers to profit and maintain their business. With zoning laws in place to

prevent land rental from increasing for agricultural sites, farmers were able to keep

production costs low and provide citizens with locally sourced fresh produce (Magigi,

2015). Zoning regulation will be a large consideration of large scale urban farming as real

estate costs rise and farmers must cope with clearing profit.

Another obstacle for any urban farmer is one of the most basic problems and one

that has developed modern agriculture into its current form: location is everything. While

a traditional field farmer may grow what best suits his or her location due to the local

climate or soil conditions, an urban farmer must create their own environment. Controlled

environment agriculture (CEA), such as greenhouse or indoor cultivation, is extremely

efficient in producing a crop with less inputs per plant than traditional means but it is

inherently very input-demanding. Creating miniature environments in an urban landscape

will require a more robust infrastructure to address the increased demand for water,

electricity, and waste disposal. Waste disposal and recycling processes exist and have

been successful in Sub-Saharan communities for organic wastes from cities to be used in

18
the food production system. Composting and sorting is utilized by the municipalities to

create a profitable recycling system of the urban wastes back into the rural agricultural

sites, creating new industries while also alleviating waste storage problems. Similar

systems could be developed in large-scale urban agriculture to reduce one of the

downsides of intensive agriculture (Drechsel, 2001). An increase in water and electrical

consumption would also be a side effect of introducing UA to a city on a large scale but

municipalities have dealt with increasing infrastructure strain since their inception.

Regulation by local governments as to the amount of UA allowed within city limits is one

approach to preventing over-use, giving utility management the time to adapt.

Distribution of the final crop is another consideration that may push investment

away from urban agriculture. DTC sales at a farmer’s market in a city or town can be

profitable for smaller growers but for realistic investment into an urban farm the market

volume must be sufficiently high and consistent. Historically, the small/local farmer must

invest their own time, often without pay, to replace the processing and distribution

network that is currently available to traditional agriculture, however, with their unpaid

efforts, they replace the “middle men”, and thus they collect a much larger share of the

retail price. This larger share comes at a hidden cost of unrecorded labor hours and higher

transportation costs per unit of produce. These hidden costs are estimated between 13 and

62% of the retail price for overhead in American produce (King et al., 2010).

The market has met these concerns with new, low-volume distributors or ‘food

hubs’ in areas where the local food is in high demand. Regional food hubs (RFH) are, at

the most basic level, a low volume distribution system for locally and regionally grown

agricultural products, ranging from livestock products to staple crops to niche-market

19
crops. RFH activities can also include marketing, advertisement, and even production

advising of the locally produced item. In a 2012 report by the USDA on RFH, they

clarify that “A regional food hub is a business or organization that actively manages the

aggregation, distribution, and marketing of source-identified food products primarily

from local and regional producers to strengthen their ability to satisfy wholesale, retail,

and institutional demand.” (Barham et. al). RFH can act as an intermediary between the

final marketplace of an item and its producer, thereby eliminating one economic barrier

to entry for the grower. In an urban agriculture setting, a RFH is especially useful as it

alleviates energy input from several different producers all bringing their crop to market

and instead resourcefully pools resources to one bulk pickup/delivery system. RFH also

acts to strengthen local and regional food systems by providing a buffer to crop failures

through redundant producers who can fill what would be otherwise empty space in a

farmer’s market or DTC sale (Low, 2015). With a strong RFH infrastructure, growers

have the opportunity to distribute more of their harvest without having to put in the extra

labor hours that would be required in DTC sale. The problem of distribution, that RFH’s

attempt to solve, appears once there is a marketable crop, and producing that crop is often

the largest barrier to entry into urban agriculture.

All new businesses that take advantage of the cutting edge of technology and the

newest consumer demands will have large upfront, capital costs, and urban agriculture is

no exception. In addition to high property values, all equipment used in a rooftop

hydroponic operation must be specifically designed and at greater expense. Thus

production values per unit area must be high, construction materials must be lightweight

and access must be provided from street level to rooftop. According to the USDA, several

20
states have begun tax incentives for individuals converting unused or run-down plots of

major cities into urban agriculture sites. With these tax incentives to new urban

agriculture ventures, the capital costs are reduced for an investor of a new urban farm

(Low, 2015). City planners can use these tax incentives to bring in new business to a city

area that may be seriously lacking an adequate supply of fresh produce and thereby gain

community involvement in a project while also eliminating part of a large barrier to

entry. Neuner, 2011, described various US projects from Alaska to the Midwest, which

required reduced capital costs for urban/local agricultural operations and offered the

potential for a healthier community that was more informed and involved with the source

of their fresh produce. While federal and local incentives may alleviate some taxes and

property costs, a large portion of the startup cost for an urban agriculture facility must

still be provided by an investor/owner. These costs are difficult to calculate for those

inexperienced in this type of business. This issue will be addressed in this project through

the development of an adaptable feasibility model for a UA site.

System Designations

The foremost concern of an urban agricultural operation is space, and the size of

the footprint required. Another concern is the weight of the system and the loads that it

will create on the building. Therefore the system should be lightweight, to avoid the

costs for strengthening the building. Creating an environment suitable for crop production

within a dense, urban area will require the installation of greenhouse or indoor growing

spaces that must be well suited for the specific climate of the site. All these factors must

be considered in the design of the growing system.

21
 There are multiple approaches to the growing system design but to reach the

efficiency demanded by the small growing spaces inherent in urban agriculture,

hydroponics must be employed. Hydroponics is a soilless culture plant growing method

which uses a nutrient solution, containing water and plant fertilizers, to grow the plants

with (or without) an artificial root zone medium for support. Hydroponics has been

proven to outcompete traditional field agriculture and is the most efficient option, even

with its inherent higher capital costs (Jensen, 1985) Closed-loop hydroponics,

recirculating of nutrients within a system, use up to 13 times less water than similar crops

grown in soil, under controlled environment conditions (Barbosa, 2015). With water

conservation becoming a larger concern of agriculture, the water-use efficiency of

hydroponics cannot be overlooked. Similarly, the nutrient fertilizer contained in the

hydroponic system is absorbed directly by the roots rather than being lost to the open

environment in soil or foliar-applied fertilizers common in field style agriculture. A

hydroponic system can be designed to optimally grow one crop. Additional benefits of

the hydroponic method contained within a controlled environment are resistance to

ambient air temperature swings, reduction of water and fertilizer inputs, and the ability to

automate operations (Jensen, 1985).

The nutrient delivery system often consists of fertilizer ‘proportioners’, or devices

which proportions concentrated nutrient stock solution at designated concentrations as

water flows through the mechanism. Proportioners allow the grower to apply a nutrient

solution at a designated concentration of water to fertilizer salts depending on the source

water and the requirements of the plants to be irrigated (Jensen, 1985). There are several

types of hydroponic systems specifically for growing lettuce, including: NFT (nutrient

22
film technique), DWC (deep water culture), and aeroponics, which are the most popular

methods. However not all three are suited for UA.

Aeroponics involves the use of high pressure misting systems within a structure

containing the roots of crops. Systems often utilize space saving designs such as a-frames

and vertical-towers and result in head densities at or above 60 heads per square meter.

However, aeroponics involve high pressure pumps which must run at intervals and have

the highest complexity of all hydroponic systems mentioned above. They also have the

highest initial capital investment and chance for system failure due to the precise nature

of the distributor nozzles which can clog easily from non-dissolved salts in the nutrient

solution. Although a very efficient growing system option, the complexity of aeroponic

systems often restricts the operations willing to invest (Christie et al., 2003).

For a small, short term (under 3 months), leafy crop, NFT can be a good choice.

An NFT system consists of gullies or channels that create rows in which the seedlings are

placed with their roots bathed by nutrient water which is pumped within the sloped

channels from a storage reservoir. The nutrient water is circulated from the storage

reservoir to each channel, either continuously or on a timed schedule, and the excess is

returned to storage. The flowing nutrient solution creates a thin ‘film’ of water as it

passes the roots (Sheikh, 2006).

The NFT approach is most well suited to the leafy greens, such as lettuce and

herbs, due in part to high oxygen availability for the partially submerged root mass, and

to provide high planting density, and easy access for frequent transplant and harvest. The

shallow nature of the nutrient stream flowing within the channel, provides a high surface

23
area to depth ratio, therefore creating high gas exchange at the film surface, increasing

dissolved oxygen in the nutrient stream, enabling more vigorous root growth.

Deep water culture (DWC), includes holding tanks filled with nutrient solution

that are topped by floating trays which support the plant at the surface of the water with

roots fully submerged and the tops above the tray. DWC can function at small levels,

with multiple, small volume tanks or in a commercial setting with long raceways, or

troughs, which run the length of a greenhouse. The raceway method allows for seedlings

to be added at one end while mature heads may be harvested at the far end, creating a

constant production line. When space or water weight is not an issue, DWC can be very

cost effective.

Despite the simplicity of DWC, it can only grow the crops in one horizontal

configuration, whereas NFT can be grown simultaneously at multiple levels, taking

advantage of the vertical space. Therefore DWC is less space efficient than high density

multi-level NFT designs. High density crop production in this document is defined as

greater than fifty heads of lettuce per square meter of growing space. NFT allows the

channels to be arranged not only horizontally, but also vertically in a three-dimensional

matrix, therefore utilizing not only the footprint of the growing area, but also the volume

above. NFT, while requiring more maintenance and technical knowledge than DWC,

does offer advantages over aeroponics in that there is a smaller chance of problems

arising in the consistency of irrigation (Graves, 1983).

High density hydroponic systems have been in use for decades as the popularity

of controlled environment agriculture grew during the 20th century. There has been

research performed evaluating specific systems and even determining their financial

24
standings. A recent study involved the creation of a novel hydroponic system which

revolved along a vertical oval pattern, dubbed a rotating living wall. Researchers

developed the system out of the necessity for a system that operated in a small footprint

but maintained the required crop production. The study focused on both the production

capacity and hypothetical financial status of an enterprise incorporating the system for

production of microgreens (Gumble et al., 2015). These studies focus solely on the

characteristics of one system in one, or a small number of scenarios.

The characteristics of the growing area are also of critical concern in urban

agriculture projects. Converting the existing structure and space into a growing area

without the major overhaul of structure is desired. There are also benefits to the existing

building, including a lower thermal load on the building, psychological benefits to the

inhabitants of the building, and even reducing noise pollution within the city (Mengual et

al., 2013). These benefits can be maximized if the proper original site is chosen that can

accommodate the weight and personnel/freight movement that will be increased from

having a production site on the roof. The roof must obviously receive ample sunlight

throughout the day but also should not be in a location that will expose it to extremely

harsh winds. The constraints involved with erecting a greenhouse or similar structure to

provide for controlled environment agriculture (CEA) must also be considered when

examining a potential space.

The greenhouse revolutionized agriculture and has now become a standard

growing method for the world’s most valuable non-commodity crops. Greenhouses with

their controlled environments and hydroponic crop production systems primarily produce

tomatoes, cucumbers, peppers, and lettuce. They can be found within many companies

25
around the world, optimal environments that will provide for the quickest and most

effective growth. There are a myriad of greenhouse structural designs each having

benefits and disadvantages, but they all are intended to maintain a stable environment for

the plants to grow and flourish, especially within areas of harsh winters and summers.

Although vital in many growing applications, greenhouses costs can be the largest factor

when considering the feasibility of a project. With the principal goal of light

transmission, followed by thermal regulation, the costs for greenhouse temperature

regulation can grow by magnitudes during the winter for heating and cooling systems

must operate almost non-stop during the summer (Jensen, 1985). Therefore, the

greenhouse glazing, or plastic outer covering, must be suited for the location and

appropriate insulation from excess heat or cold must be included.

Objectives

The objective of this research was the development of an adaptable cost of

production computer model which may be applied to almost any urban area and provide a

basic analysis of the economic feasibility of a specific production site. This was

accomplished through use of proven capabilities of the novel hydroponic system as a

basis for calculating the production parameters of a hypothetical urban agriculture

operation. The secondary objectives of the study were to design and construct a high

density hydroponic system with appropriate design characteristics for a rooftop, urban

agriculture environment; and, to verify its operation and its capability to produce a

marketable lettuce crop within a high density hydroponic nutrient film technique system.

26
 MATERIALS AND METHODS

A-Frame Design and Construction

Research to design and construct a high density hydroponic system was

conducted at the University of Arizona Controlled Environment Agriculture Center from

January 2015 through May 2015. This system was planned for use in the urban

agriculture environment, specifically on the rooftops of existing structures.

The hydroponic system was designed and constructed at the ABE

fabrication shop at the Campus Agricultural Center with the aid of Charlie Defer, senior

fabricator, and Don Clifford. The system would provide the highest growing density in

the smallest footprint by utilizing vertical space while also remaining as lightweight as

possible to avoid roof reinforcement costs. An ‘A-Frame’ style system employing the

Nutrient Film Technique (NFT) was chosen. This hydroponic method allows for

maximization of incoming sunlight through sloped sidewalls, while also inherently

creating less weight from the small water reservoir. The utilization of the three

dimensional space provides greater plant density than other horizontal hydroponic

methods, such as floating raft or NFT without an A-Frame.

The design of the A-frame system was comprised of three aluminum trusses.

Each truss was constructed of two aluminum struts, each 213 cm, and connected at one

end to form an ‘A’-shaped truss. Sixteen 3.81 cm diameter x 305 cm long schedule 40

PVC pipes were attached to each face of the three trusses, to form the three-dimensional

A-Frame structure that held 32 plant rows, as seen in Figures 1 and 2. The PVC pipes

were connected to the faces of the trusses within the A-frame with aluminum conduit

straps and self-tapping screws, and the PVC pipes and A-shaped trusses formed a solid

27
frame. The structure had a footprint of 1.5 x 3.05 m for a total of 6.5 m2. The PVC pipes

were each a row of plants with 2.5cm diameter holes, spaced 15 cm apart for a total of 20

holes per row. The plant row pipes were mounted to provide a 15 x 15cm plant spacing

as viewed from overhead.

The plumbing manifold for nutrient water outflow and return was attached to the

A-frame, and as indicated in Figure 4. The 416 liter nutrient water reservoir was located

beneath the A-frame which provided shade for the tank, reducing algae growth and

heating of the nutrient solution, and to effectively utilize the empty space Figure 3. The

A-frame was oriented north to south, having an east face and a west face for the plants.

28
Figure 1. A computer rendering of the a-frame design using Solid Works software.

Figure 2. A view of the East face of the a-frame hydroponic system. Note the holes along
the length of each pipe and the first-generation reservoir, later upgraded to a larger tank.

29
Figure 3. A view of the South face of the a-frame system still under construction,
highlighting the aluminum struts forming the support structure of the system which all
components are attached to.

Figure 4. The Northern face of the system, the reservoir of the system is placed within the
empty space created by the a-frame design. The 416 liter reservoir is insulated with a
solar wrap and Styrofoam blocks to limit solar heat gain and to prevent algae growth
within the tank.

The plumbing manifold delivered the nutrient solution to the thirty two plant rows

with an external circulation pump (Little Giant Pump Co., Oklahoma City, OK) that was

plumbed into a 1.9 cm diameter bulkhead fitting in the end of the nutrient water reservoir.

The external circulation pump was sized through calculation of total resistance in the

system, 3.29 meters and compared to performance curves. A Little Giant ‘3-md-sc’

model was chosen considering the pump capacity is approximately 35 lpm at a head of

30
3.3 meters and the required flow to the system was calculated to be ~32 lpm. The

calculation procedure for the pump sizing can be seen in appendix item iii. The nutrient

solution flowed from the circulation pump at approximately 64 lpm up a vertical riser

through an inline filter and then split into two manifolds at the peak apex of the A-frame,

and then to both of the legs of the southern-facing aluminum trusses. Along the length of

the 13 mm diameter manifold pipes were four even spaced pressure regulating flow

splitters (Azusa, CA). From each of the splitters, four, 6.5 mm diameter flexible plastic

tubes were connected to each of four adjacent plant rows. The inlet manifold along the

South face includes a pressure by-pass valve for directing some of the pump flow directly

back to the nutrient storage tank and provide flow control for the combined system of

plant rows. The flow at each plant row was between 1-2 lpm.

The 3.81cm diameter schedule 40 PVC return line conveyed the nutrient solution

drainage from each plant row, after passing through the 20 plant roots contained in the

NFT channel, to the reservoir for re-circulation. The A-frame system was situated at a

1:100 elevation loss to allow the nutrient solution to flow from the input manifold,

through the tubes, passing the roots before finally returning to the reservoir. The return

plumbing was sanitary tees connected individually at the end of each of the plant row

PVC pipes which then formed a column extending down the A-frame on the North end.

At the bottom the tees trace into one pipeline and eventually flow into a nutrient storage

tank (Figure 5). Both the return line and inlet manifold have unions and a ball valve to

allow for easy maintenance. The entire system was raised 15 cm to provide room

underneath the lowest NFT channel for the return line to have sufficient head into the

system without backup. The additional height was added due to the second generation

31
reservoir having a taller upper rim, a trough-style reservoir would be optimal for this

system.

Figure 5. View of the return line bulkhead into the reservoir from the North face, a union
was placed in-line to allow for simple servicing.

Crop Trials

Crop trials were completed at the University of Arizona Controlled Environment

Agriculture Center from September through December 2015. The crop response,

electrical power and nutrient solution consumption were recorded to quantify

32
performance characteristics of the system. Environmental conditions, including air temp,

PPF, pH, EC, others were monitored…

The experiments were completed within an arch-style greenhouse located at the

University of Arizona Controlled Environment Agriculture Center. It is a single bay

greenhouse with a footprint of 9.2 x 14.6 m and an apex height of 4.9m. The greenhouse

is equipped with air-inflated double layer polyethylene glazing film and 8 mm, acrylic-

coated polycarbonate rigid plastic end-walls. The greenhouse utilizes a fan and pad

evaporative cooling system consisting of two 3.4 m3 s-1 exhaust fans and an 8.5 m x 1.5m

pad. The North-South oriented structure contains an EnviroSTEP environmental

controller which monitors air temperature and relative humidity inside an aspirated

container (Wadsworth Control Systems, Inc., Arvada, CO). The greenhouse also featured

a natural gas-fed hot-air heater located above the evaporative cooling pad on the Northern

wall. The A-frame system was located in the Southwest corner of the greenhouse.

The trials tested two factors, including: that the system produced a crop of equal

size and quality on both the East and West faces; and, that high density spacing (70+

heads m-2) would not diminish growth in comparison to low density spacing (35+ heads

m-2).

All seedlings were started in Jiffy ‘Preforma’ commercial seedling cubes (Figure

6), which was a mixture of fine peat and a binding agent and has been widely used starter

plug in the agricultural industry. The seedlings, each measuring 3.81 x 2.5 x 1.9 cm, were

prepared in perforated polystyrene trays measuring 51 x 25 cm, consisting of 105 pre-

formed cutouts, with drainage. The seedlings were prepared at a depth of 2 cm into the

33
peat material in crops of 105 seedlings per tray. Once seeded and imbibed with untreated

tap water, the trays were placed into the environmental growth chamber located at the

University of Arizona’s Controlled Environment Agriculture Center. . The growth

chamber (Figure 7) air temperature was 18°C for the first three days, then raised to a final

temperature of 24°C prior to transplant in the greenhouse. A 295 watt light emitting

plasma-based growth light (Stray Light Optical Technologies, Inc.) provided

approximately 14,000 lumens at a height of ten inches. The photoperiod selected for

germination and seedling maturation was adopted from the Cornell CEA Handbook

(Brechner, 2001) and consisted of a twenty-four hour photoperiod for the first eleven

days of seedling growth followed by a fourteen hour photoperiod for the remaining three

days. Non-fertilized municipal tap water was manually irrigated over the seedling trays

once per day using a pump sprayer; a total of 7.57 liters of water were used from the

planting of the seedlings until their eventual transplant.

34
Figure 6. An example of a Rex lettuce seedling grown in Jiffy Preforma material, nine
day growth from start of seeding.

Figure 7. The Jiffy Preforma growing material trays were germinated and grown to an
age of 14 days inside the environmental growth chamber at the University of Arizona
Controlled Environment Agriculture Center.

35
Figure 8. The seedlings were placed at a height of 45.7 centimeters below the light fixture
after the initial twenty four hour period inside the growth chamber. What was the PPF?

 First Trial: Small batch of 67 heads, testing general operation of system,

9/3/15 thru 10/15/15

 Second Trial: 165 heads in system, tested equal growth among East/West

faces, 9/17/15 thru 11/7/15

 Third Trial: 176 heads in system, increasing density for fourth trial,

10/4/15 thru 11/29/15

 Fourth trial: 205 heads in system, tested ability for rows to be at full

capacity, 10/22/15 thru 12/15/15

A crop layout was constructed using Microsoft excel to outline the transplant

process and describe placement of heads during each of the four experiments. An

experiment is defined as the individual crop trials seeded and transplanted into the

36
system, consisting of two Jiffy trays each. The experiments overlapped in time as the A-

frame was never completely filled, and there would be two experiments in parallel in the

system during a majority of the time from September to December.

The crop layout document consisted of five timeframes, from the first day after

the first experiment had been transplanted in up until the harvest day of the fourth

experiment. Each timeframe represents one of the five vertical areas in the excel sheet

consisting of one portion stating the date set, one portion describing the layout on the

West side and one portion describing the layout on the East side, moving from top to

bottom. Within each timeframe, the layout is partitioned the system into East and West

faces and further into the rows or tube numbers that make up that face. The layout

document was used to accurately test the specific questions surrounding the system with

clarity to the researcher and allowed the research to be statistically sound through proper

research methods by eliminating possibly confounding variables through confirmation of

equal means between faces and spacing layouts of the system.

Transplant of the seedlings to the A-frame system was done fourteen days from

the day of seeding, after five in the afternoon to prevent stress to the plants from intense

sunlight. According to the crop layout for the specific crop cycle, the seedlings were

arranged in either single or ‘every-other arrangement’, i.e., occupying all grow spots in a

plant row or occupying every other linear spot. The seedlings were placed into the 2.5 cm

diameter holes with 1.3 cm of the grow-cube above the tube and the lower portion

situated inside the tube to allow root hydration and prevent excess light entering the tube.

Prior to the new transplants, the NFT channels to be used were cleaned and all manifold

equipment was flushed. The contents of the reservoir were also examined and the nutrient

37
solution parameters were measured to affirm that the system was ready to receive the new

heads.

The parameters of the interior greenhouse climate and nutrient solution were

measured using Hanna stationary meters (Hanna Instruments Inc., Rhode Island) or with

a CR23x-4m datalogger system (Campbell Scientific Inc., Logan, UT). Measurements

were sampled every 10 s and recorded to the datalogger every fifteen minutes for the

duration of the experiment. Climate parameters measured included: air temperature (°C),

relative humidity (%) and interior PPF (mol m-2 s-1); and, reservoir measurements

included: dissolved oxygen content (ppm), electrical conductivity (mS/cm), and pH. The

data were downloaded weekly and imported into Excel to be analyzed. The datalogger

system accurately determined trends in the aerial climate and the root zone for consistent

plant growth.

Four crop harvests were completed during from Sept to Dec 2015 to test specific

transplant layouts and to determine quantifiable growth among the lettuce. Ten heads

were harvested from each selected row and weighed using. They were then placed into a

drying oven for 7 days at 50°C and weighed using an Adventurer analytical scale (Ohaus

Corp., Parsippany, NJ) .The first two experiments established that the system could run

consistently and, above all, that the east and west faces of the A-frame would produce

similar heads and therefore not become a variable in itself. Therefore, the second

experiment was transplanted in the same layout on both faces (East and West). The third

experiment was completed to both fill the system to a greater capacity and to form a gap

between the second and third experiment. The fourth and final experiment evaluated crop

38
response to multiple rows (more than three consecutive rows) and how they may appear

in an actual, commercial production, meaning that the a-frame is filled to full capacity.

Confirmation of the validity of the system was done using statistical analysis

software SAS (SAS Institute Inc., Cary, NC). ANOVA analyses were performed

assuming an alpha, or the probability of rejecting the null hypothesis i.e. believing a

‘false positive’], of (α = 0.05) and testing a null hypothesis that within specified crop

trials there would be no significant statistical difference between the two planting layouts

that were being tested. The experiment was completed to test the growth on the East and

West faces of the system in the second trial and to test for deficit in growth from

cropping multiple rows as single spaced (SS) or full capacity, in comparison to multiple

rows in every other (EO), 50% of the row was planted, format. With confirmation of

success in the hydroponic system and an effective transplant layout determined, the

research shifted to developing the economic investment model.

Economic Model

With the establishment of the A-frame system as a viable growing structure, an

economic model of a hypothetical growing operation was initiated. The basis for the

model is a simple investment decision aid for an urban agriculture operation that can be

applied to various locations, from a limited amount of information inputted.

Introduction

39
 The economic model was built for an investor from most any profession to

understand while also providing the experienced grower a robust method for developing a

basic cost-analysis of a UA operation. The grower will also input other factors that relate

to the greenhouse structure such as weather factors and cooling usage. The model relies

on a user-determined number of “crop cycles” per year, or, the number of crop harvests a

grower can accomplish within one calendar year. The model displays a financial output

spread over a user-determined set of years to alleviate high upfront capital expenditures

and to not skew the model for the short term return on investment of fixed costs.

The model has a ‘main’ sheet, which clearly displays inputs, costs, and revenue

analysis from several ‘auxiliary’ sheets. The auxiliary sheets include:

 Heating

 Cooling

 Labor

 Additional Materials

 Nutrient Solution Costs

 A-frame build cost

These sheets which supply their calculated results into the main sheet, form the final cost-

analysis.

Heating

The heating sheet is the most extensive auxiliary sheet and requires the user to

input local climate data for their operation. Using an empirical formula for calculating

40
heat loss from a greenhouse, through the input of climate temperature, set point air

temperature inside the greenhouse, and greenhouse construction materials (Eq 1), the

energy loss can be calculated each month and summed for the year (Aldrich and Bartok,

1994).

𝑄 = 𝑈 𝑥 𝑆𝐴 𝑥 (𝑇𝑖𝑛,𝑚𝑖𝑛 − 𝑇𝑜𝑢𝑡,𝑚𝑖𝑛 ) Eq 1
where U =Thermal quality of glazing
SA =Surface area of greenhouse
Tin = Set point temperature inside greenhouse
Tout =Average temperature outside

Labor Sheet

Labor is another auxiliary sheet which uses a per-operation style approach to

calculating labor. The approach factors in time spent per activity per crop item, including:

 Delivery

 Marketing

 Seed/Transplant/Harvest/Package

 Production Management

 Maintenance

The labor costs were based on the number of plants involved in each cycle and the

number of cycles per year. The user must provide the hourly labor rate and the time

required for each task. The time required per task can be determined using direct time

study, a method of motion analysis where highly repetitive tasks can be assigned

definitive time periods. (Luxhoj and Giacomelli, 1990). With the total number of hours

involved per cycle/year calculated, the sheet applies the hourly rate to output a total labor

cost in terms of cycle and year.

41
Nutrient Solution Sheet

The nutrient solution auxiliary sheet calculates the cost per liter of nutrient

solution, based on nutrient formulation, and excluding municipal water costs which

appear later on the Main sheet. The nutrient solution sheet provides a list of chemicals

and their proportions required for a general formulation of a modified Hoagland solution

(Hoagland and Arnon, 1950) (See appendix). It includes a cost for each chemical and the

calculation of total costs for the maximum use of each stock tank and the cost for a

system nutrient reservoir per cycle.

A-Frame Build Sheet

The A-frame build cost sheet outlines all parts necessary for the construction of

the A-frame hydroponic system which forms the basis of the financial model. (See

appendix) The total build cost for one hydroponic A-frame system is documented and is

adjustable to the user based on individual pricing and build preferences.

Cooling Electrical Sheet

Cooling and miscellaneous electrical costs are included in the Cooling Electrical

auxiliary sheet. This sheet is used to calculate electrical costs for cooling with ventilation

fans while also including an hourly use (kWh) of electrical demands other than exhaust

fans. The additional electrical input resulted from the electrical consumption of the

nutrient delivery pump of the A-frame system. The irrigation cyclic usage based off

number of a-frame systems at the operation and the six week usage, about 230 kWh, of

42
the pump. Calculation of exhaust fan capacity required was based on 2.4 m3 of air flow

per m2 of growing space (Aldrich and Bartok, 1994). This sheet uses the inputted

growing space from the main page and with the user input of exhaust fan yearly usage,

and miscellaneous electrical usage per cycle to provide an output of the total electrical

cost per cycle and per year.

Additional Materials Sheet

The Additional Materials sheet was to be utilized by the grower as an area to

include specialty items that are turned over yearly and must be replaced, such as

packaging materials, safety equipment, and backup equipment. The Additional Materials

sheet provides these additional costs into the final cost breakdown at the conclusion of

the Main page.

Main Page

The Main page of the Excel document combines the values of all the auxiliary

sheets along with the production capacity of the system and a cost analysis. The main

page is split into five distinct areas:

 Customer Inputs

 Assumptions of the A-frame system

 Facility Designations

 Crop per cycle/year

 Cost Breakdown

43
 The user inputs were made to be applicable to potential investors from non-

agricultural backgrounds as well as growers with extensive experience. There are sixteen

inputs to the model:

 Available growing space

 Utility costs

 Growing material costs

 Land leasing cost

 Grow cycles per year

 Greenhouse thermal quality

 Greenhouse Surface Area

 Observation period (lifetime of equipment)

 Media costs

 Nutrient solution cost

 Hydroponic system cost to build

 Labor pay rate

 Heating fuel price

 Greenhouse construction cost

 Market price per head

All but two inputs, ‘cost per liter of nutrient solution’ and ‘cost per system’, are

available to the grower through the local gas, water/sewer, electrical utility providers, and

the local market pricing. The input area of the main sheet is shown in Figure 9.

44
 Figure 9. An example layout of the input section to the main sheet.

The descriptive factors about the hydroponic system and the greenhouse are

described in the ‘Assumptions of the System’ portion of the Main page. This section

serves to clarify the source of numbers used in later calculations throughout the

spreadsheet. The factors for this area come from the crop trial experiments that were

performed. These values may be changed by the user if they would like to compare other

growing systems with different capabilities but they are arranged for use in a model

incorporating the A-frame hydroponic system. The assumptions for this system include:

 A-Frame Footprint

 Seedlings per Tray

 Water Use per System, per Cycle

 Cycle Duration

 A-Frame Capacity

 Over-Seeding Correction

 Assumed Crop Survival

45
  Total Number of Cycles Observed

The assumptions of the A-frame system are displayed in Figure 10

Figure 10. The assumptions of the system as they appear on the Main sheet.

The Facility Designations portion of the Main sheet is derived from previous

inputs and other auxiliary sheets. Figure 11 shows the four outputs of the Facility

Designations area. This allows for a perspective on the scale of the operation and it

includes information such as:

 Number of A-frame Systems

 Exhaust Fans Required

 Labor per Cycle

 Cost for All Systems

Figure 11. Facility designations of the model which are calculated outputs of previous
inputs, on the main sheet.

46
 The Crop Output portion of the Main Page displays the current crop requirements

and maximum output to the user after all data inputs are provided. The crop is partitioned

into both a cycle and total yearly analysis. The growing media, including seed and trays,

is determined from the assumptions of the number of seedlings a seedling tray can grow

(ie. capacity of the seedling tray), and the number of heads the facility can produce in a

cycle.

The fertigation requirement is calculated from the number of A-frame systems

that are within the Facility Designations and the Assumptions Section regarding water

use per system. The maximum number of heads for each cycle and per year is determined

by the number of heads produced per system and is multiplied by the crop survival rate,

typically between 80 and 90% in CEA hydroponics, in the Assumptions of the System

section. Figure 12 displays an example production of the crop area of the Main Sheet

with an example grow space inputted.

Figure 12. The display of the crop section after calculating a given set of inputs.

The cost breakdown is the final result of the model and crop trials. It displays the

variable and fixed costs of a hypothetical urban agriculture operation including the

revenue stream, ROI (Return on Investment)/first year investment, sensitivity of yield

47
analysis, and net cash position. The variable costs fluctuate with and are directly related

to each production cycle, while the fixed costs remain constant through the observed

period. Figure 13 is a hypothetical output of the variable and fixed costs portion of the

cost breakdown. The input to the main sheet of ‘market price per head’ is the largest

determining factor in the cost breakdown and allows the user to specify local market

conditions and ultimately, determine the feasibility of a site for an urban agriculture

facility.

Figure 13. A typical output of the variable and fixed costs portion of the cost breakdown

on the main sheet.

A Year 1 investment analysis was constructed by determining the total capital

involvement to a hypothetical area and performing a Return on Investment (ROI)

calculation as a percentage of total capital costs from the first year including permanent

structures and materials. The total investment analysis is important to all investors to

determine the investment risks and rewards of a potential site and give this model more
48
credence to its validity and accuracy. Along with an ROI analysis, net cash position

graphs were created in both cycle and yearly view. Figure 14 shows the sensitivity

analysis of a hypothetical UA operation. The Sensitivity analysis provides the user with

an instant readout of the best/worst case scenario of multiple situations at once through a

data table calculation. The data table runs the model for various inputs along an x and y

axis orientation that give a profit output. This output shows the profit that may be

obtained at various yields and market prices.

Figure 14. An example sensitivity analysis, note the yield percentages along the y axis,
the market prices along the x axis, and the profit-per-cycle as the response.

The net cash position graph is a visual of the capital investment throughout time of the

operation. The graph, features a monetary y axis with time (in cycles and years) along the

x axis. As the hypothetical operation moves through time, starting at the negative cash

point of total investment, profits each cycle or year allow the cash point to move to and

above the x axis, indicating the business is operating at a profit. Through subtraction of

total returns less variable costs from the total investment sum, graphs show the break-

even point of the operation over the observed period of time

49
 RESULTS and DISCUSSION

A-Frame Design

The A-frame hydroponic system performed with the only loss occurring due to

manifold clogging. The clogging of the manifold was solved with two additions to the

system: an inline 50 mesh micron filter placed between the circulation pump and the apex

of the riser and more robust pressure-regulating irrigation manifolds. The micron filter

screen was cleaned daily to prevent particulates from entering the nutrient solution

delivery plumbing. In Figure 15 the addition of the inline filter, the pressure regulating

emitters, and the flow-reducing valve can be observed.

Figure 15. A view of the Eastern face of the system, the inline filter (A) can be seen along
the main riser, above the pressure relief valve (C) and pump. One may also note the

50
pressure-regulating emitters (B) along the diagonal aluminum frame of the system with
individual feeder tubes directed into individual channels.

The first experiment identified the size of the reservoir as a critical point of the

system. The reservoir must be as large as possible to allow for the consumption of

nutrient water. Assuming that there was no automated filling mechanism, daily attention

to this was necessary. There was 1.04 L of nutrient water in storage per plant. Peak

water consumption was 45 L day-1. Furthermore, as nutrients were consumed, the pH of

the nutrient storage water increased to 6.8. . As heads of lettuce grow and absorb nitrate

ions into the root system they release a base hydroxyl ion to maintain a neutral level

within the roots. This process is usually done at such a rate that the buffering capacity of

a hydroponic system’s reservoir can maintain a stable pH within the recommended 5.5-

6.5 range. However, in a system with such a high ratio of plants to volume of nutrient

solution, the release of the base hydroxyl ions can cause large shifts in the pH of the

reservoir if water storage volume decreases. At a pH greater than 6.5, certain nutrients

become unavailable for absorption by the plants, as shown by Figure 16, causing adverse

plant growth effects which can reduce head fresh weight, diminish appearance of green

color, and taste. Therefore, the system was refilled daily with fresh nutrient solution, at a

slightly more acidic pH of 5.2, and the reservoir was enlarged from 265 liter to 416 liters,

therefore increasing the value to 0.7 L per head of lettuce from 1.04 L per head. An

automated refill system was considered but ultimately decided against considering daily

observations of the system were being conducted regardless of the level insides the

reservoir.

51
Figure 16. The absorption of macro and micro nutrients varies with pH of the solution,
particularly, several micro nutrients such as iron and manganese become nearly at a pH
above 7.0. (Kujawski, 2014)

The A-frame system was designed with relatively small 3.81 cm inside diameter

PVC pipes, to maximize diagonal space along the A-frame and to minimize shading to

the rows of plants beneath each pipe. While highly space-efficient, the narrow diameter

presented a problem with the growth of the root mass and required the addition of a

bypass valve in the main riser, located immediately after the outlet of the pump. The

bypass valve consisted of a plumbing tee from the main riser which directed the pump

output flow back into the reservoir. It was controlled by a union ball valve. The gate

valve is opened as the plants grew to maturity, allowing for a decrease in nutrient water

flow as the root mass increased in the pipe and blocked the flow, and reduced over-

flooding of the rows but maintained a thin film of nutrient solution flowing within all of

the plant rows.

52
 The final design of the A-frame system had a maximum capacity of 608 heads

with a 4.65 m2 footprint. Incorporating a 30.5 cm ‘work area’ around the entire perimeter

of the A-frame footprint, the hydroponic system occupied 7.8 m2 and a 2.1 m maximum

height at the apex. Including the workspace required for access around each A-frame, the

system produced 77 plants m-2 which is 44% greater than traditional horizontal growing

systems with 43 plants m-2 (Hernandez, et al, 2016)

Crop Trials

To quantify the novel system and utilize realistic production values in the

economic model, four production trials were performed in the fall of 2015.

First Experiment

The first experiment provided a harvest from four rows of the system. It was

completed as a preliminary trial, and it identified the clogging and reservoir issues within

the system quickly and allowed the system to be modified and adapted. The layout for the

first experiment was according to the crop layout document.

Table 1 is the harvest weights of individual rows and the respective analysis. The

first experiment showed relatively uniform growth among the four rows, excluding E7

which experienced a technical clogging failure. Table 2 shows the dry weight analysis of

the first experiment and shows less uniformity in average head weight among rows. The

graphical representation of the fresh weights and dry weights of the first experiment

harvest are shown in Figure 17 & 18, respectively. All improvements to the system were

53
completed during the first experiment and remained unchanged for the remaining

experiments.

Table 1. Fresh weight of the first experiment harvest. Rows are represented by face (W=West,
E=East) and by number (1 to 16, from bottom to apex). Total head mass (g), Total root mass (g),
Average head mass (g) of ten sample size, Average root mass per head (g), Root mass
percentage (%) for each of two rows from East side face (E1 & E7), and two rows from West side
face (W1 & W6) of the A-frame.

Fresh Weight W1 W6 E1 E7

Total Head Mass per Row (g) 2135 2260 1955 1010

Total Root Mass (g) 530 565 810 430

Average Head Mass (g) 213.5 226.0 195.5 101.0

Average Root Mass per Head (g) 53.0 29.7 42.6 22.6

Root Mass Percentage (%) 25 13 22 22

Table 2. Dry weight of the first experiment harvest. Rows are represented by face
(W=West, E=East) and by number (1 to 16, from bottom to apex). Total head mass (g),
Total root mass (g), Average head mass (g) of ten sample size, Average root mass per
head (g), Root mass percentage (%) for each of two rows from East side face (E1 & E7),
and two rows from West side face (W1 & W6) of the A-frame.

Dry Weight W1 W6 E1 E7

Total Head Mass / Row (g) 70.9 56.4 82.7 38.9

Total Root Mass (g) 30.4 33.8 32.9 34.5

Average Head Mass (g) 23.6 18.8 27.6 13.0

Average Root Mass per Head (g) 3.00 1.80 1.70 1.80

Root Mass Percentage 13% 9% 6% 14%

54
 400
350
300

Head Weight (g)

250
200
150
100
50
0
W1 W6 E1 E7
?Samples within each Row? Row/ Individual Sample

1 2 3 4 5 6 7 8 9 10

Figure 17. First experiment harvest fresh weights. Each bar represents one of the ten samples
from each of four rows.

45
40
35
Head Weight (g)

30
25
20
15
10
5
0
W1 W6 E1 E7
Row/ Individual Sample

3 5 7

Figure 18. Lettuce Head average dry weight for first experiment. Each bar represents one of the
ten samples taken from each row, each grouping of bars represents the weights of a whole row.

Second Experiment

The second experiment included the improved design and operation of the A-

frame system, and an increased number of plants from 67 to 165, raising the system

capacity from 11% to 38%

55
 With the improvements established from the first experiment, a greater number

of plants were successfully grown, having less root zone growing material entering the

nutrient stream (use of an inline filter), more uniform flow rate to each row (pressure

compensated emitters) and maintaining the nutrient solution pH at desired value (larger

reservoir volume and daily addition of fresh water).

The purpose of the second experiment was to demonstrate the uniformity of plant

production and crop harvest between the east and west faces of the A-frame. The

seedlings were transplanted within nearly identical layouts, (Figure 19) of the heads

within and among the rows of each face of the A-frame. There were 12 rows and 10 or 19

plants within each row, depending on spacing pattern designation. An ANOVA statistical

analysis (SAS) provided verification that there were no significant statistical differences

at 95% confidence between the heads grown on either face. Both ANOVA analyses

exhibited p values of 0.85 and 0.30 for the fresh and dry weights, respectively. Figures 20

and 21 are graphical representations of the fresh and dry weights from the second

experiment. They indicate a more uniform average head fresh weight across all the rows

with a standard deviation under 25 grams for a population of 165 heads. Tables 3 and 4

represent the fresh and dry weights of the second harvest, respectively.

Table 3. Total Head and Root Fresh Mass, and Average Head and Root Fresh Mass for 10 plants
in the second experiment harvest. Calculations were done in the same fashion as described in
table 1.

56
Table 4. Total Head and Root Dry Mass, and Average Head and Root Dry Mass for 10 plants in
the second experiment harvest. Calculations were done in the same fashion as described in
table 1
Dry Weight W2 W4 W5 W8 w9 w10 E2 e4 e5 e8 e9 e10
Total Head
Mass / Row
(g) 44 44.2 45.7 45.2 47.6 42.1 46.8 46.02 52 47.6 43.1 43.9
Total Root
Mass (g) 18.9 21.4 26.3 21.7 19.0 20.3 22.0 20.4 30.5 17.9 16.5 18.8
Avg. Head
Mass (g) 14.7 14.7 15.2 15.1 15.9 14.0 15.6 15.3 17.3 15.9 14.4 14.6
Average Root
Mass per
Head (g) 1.0 1.1 1.4 2.2 1.9 2.0 2.2 1.1 1.6 1.8 1.7 1.9
Root Mass
Percentage 7% 8% 9% 14% 12% 14% 14% 7% 9% 11% 11% 13%

57
 Layout from 9/29/15 thru 10/15/15 (1st + 2nd Crop)
Side Row Arrangement Crop # Heads Needed

West 1 Every Other 1 10

West 2 Single 2 19
West 3
West 4 Single 2 19
West 5 Single 2 19
West 6 Single 1 19
West 7
West 8 Every Other 2 10
West 9 Every Other 2 10
West 10 Every Other 2 10
West 11
West 12
West 13
West 14
West 15
West 16

East 1 Single 1 19
East 2 Every Other 2 10
East 3
East 4 Single 2 19
East 5 Single 2 19
East 6
East 7 Single 1 19
East 8 Every Other 2 10
East 9 Every Other 2 10
East 10 Every Other 2 10
East 11
East 12
East 13
East 14
East 15
East 16
Total Heads 232
Percent Full 38%

Figure 19. All heads under the designation of ‘crop #’ that are labeled with a ‘2’ were from the
second experiment and their placement is designated in the above excerpt from the crop layout
document. The spacing style (Single or Every Other) and the subsequently required number of
heads (10 or 19) is also specified.

350

300

250
Head Weight (g)

200

150

100

50

0
W2 W4 W5 W8 w9 w10 E2 e4 e5 e8 e9 e10
Row/ Individual Sample

1 2 3 4 5 6 7 8 9 10

58
Figure 20. Fresh lettuce head weight for 10 samples (1 – 10) from the second experiment for
each row number (2 through 10), and each face (East (E) or West (W)).The colors within each
grouping represents the different rows.

20
18
16
14
Head Weight (g)

12
10
8
6
4
2
0
W2 W4 W5 W8 w9 w10 E2 e4 e5 e8 e9 e10
Row/ Individual Sample

3 5 7

Figure 21. Dry lettuce head weight for 10 samples (1 – 10) from the second experiment for
each row number (2 through 10), and each face (East (E) or West (W)). The colors within each
group represents the different rows.

Third Experiment

The third experiment tested greater plant density and raised the system production

capacity from 38% to 63%, as shown in the crop layout document, by increasing the

number of plants from 165 to 176.

Harvest for the third experiment was delayed 7 days compared to the Second

Experiment to study increased head size and overcrowding between plants to determine if

a larger head size would be attainable in the current system. No detriment to plant growth

from factors such as mold, abnormal shoot growth, or root growth problems were

observed. However, by allowing the plants to grow beyond the market weight of 150

59
grams, did cause problems with overflow of nutrient solution from the PVC grow tubes

and subsequent loss of solution and flooding.

With the delay of harvest, the lettuce head fresh weights were more inconsistent

between rows than previous experiments, with a standard deviation among all rows of 70

grams for 176 heads compared to 85 for previous. A delayed harvest, was applied to the

system, the greater head mass and higher variation among rows was seen as useful data

and could be used in future research to adapt the system to larger grow cycles. Figures 22

and 23 display this difference of growth among rows graphically and in comparison to

figures 20 and 21, show the greater standard deviation. Tables 5 and 6 are the fresh and

dry weights of the third experiment harvest, respectively.

600

500
Head Weight (g)

400

300

200

100

0
W7 W11 W12 W13 W14 W15 W16 E7 E11 E12 E13 E14 E15 E16
Row/ Individual Sample

1 2 3 4 5 6 7 8 9 10

Figure 22. Fresh lettuce head weight for 10 samples (1 – 10) from the third experiment
for each row number (7 through 16), and each face (East (E) or West (W)).The colors
within each grouping represents the different rows.

60
 30

25

Head Weight (g)

20

15

10

0
W7 W11 W12 W13 W14 W15 W16 E7 E11 E12 E13 E14 E15 E16
Row/ Individual Sample

3 5 7

Figure 23. Dry lettuce head weight for 10 samples (1 – 10) from the third experiment for
each row number (7 through 16), and each face (East (E) or West (W)). The colors within
each group represents the different rows.

Table 5. Total Head and Root Fresh Mass, and Average Head and Root Fresh Mass for
10 plants in the third experiment harvest. Calculations were done in the same fashion as
described in table 1.
Fresh
Weight W7 W11 W12 W13 W14 W15 W16 E7 E11 E12 E13 E14 E15 E16

Total Head
Mass /
Row (g) 4315 3095 3340 2555 2990 2620 2890 3780 3555 2425 2045 1865 1790 2585
Total Root
Mass (g) 567 451 620 320 411 310 500 461 435 326 455 330 328 533
Average
Head Mass
(g) 432 310 334 256 299 262 289 378 356 243 205 187 179 259
Average
Root Mass
per Head
(g) 56.7 45.1 32.6 32.0 41.1 31.0 26.3 46.1 43.5 32.6 23.9 33.0 32.8 28.1
Root Mass
Percentage 13% 15% 10% 13% 14% 12% 9% 12% 12% 13% 12% 18% 18% 11%

Table 6. Total Head and Root Dry Mass, and Average Head and Root Dry Mass for 10 plants in
the third experiment harvest. Calculations were done in the same fashion as described in table 1

61
Dry Weight W7 W11 W12 W13 W14 W15 W16 E7 E11 E12 E13 E14 E15 E16
Total Head
Mass /
Row (g) 70.5 57.1 60.9 48.6 52.5 55.4 61.3 71.6 62.3 49.5 50.9 44.4 46.5 54.8
Total Root
Mass (g) 34.7 29.3 41.1 25.9 31.9 28.8 34.3 31.9 26.9 26.2 54.1 33.2 26.9 34
Avg. Head
Mass (g) 24 19 20 16 18 18 20 24 21 17 17 15 16 18
Average
Root Mass
per Head
(g) 3.5 2.9 2.2 2.6 3.2 2.9 1.8 3.2 2.7 2.6 2.8 3.3 2.7 1.8

Root Mass
Percentage 15% 15% 11% 16% 18% 16% 9% 13% 13% 16% 17% 22% 17% 10%

Fourth Experiment

The fourth experiment was planted at the highest density (87 Heads m-2) and it

was performed to test for any detriment to growth that would occur from running the

system at full capacity. The East face was planted 100% filled with 19 plants per row for

rows two through six while the West face was planted at 50% or 10 plants per row for

rows two through six. The 50% occupancy was accomplished by planting every other

hole along each row with the remaining holes left empty.

This highest density planting yielded no differences among plants rows on either

east or west face. This experiment required that the bypass valve divert water back to the

nutrient reservoir and to reduce the total output from the pump and the flow rate to each

row, which provided a reasonable flow and level of water within the row tube, and

prevented spill-over from the root mass blocking nutrient solution flow along the tube.

The bypass valve was first opened to 25% reduction of flow on the fourteenth day after

62
transplant, and was further opened to 40% on the 22nd and another 10% again on the 28th

day, ending one-half closed by the day of harvest.

Due to a reduction in incoming sunlight due to the Winter months, the plants in

the fourth experiment grew at a slower rate, requiring nearly the same time period to

reach an acceptable market weight of 150 grams that the third trial required to reach fresh

weights over 300 grams. This slower growth rate was attributed to the environmental

conditions, incoming solar radiation, and not to the spacing arrangement. Figure 24

graphically displays the fresh weights of the heads in the fourth experiment with a

standard deviation among means of 22 grams, among the 150 heads that were grown in

the arrangement patterns. Tables 7 and 8 were the fresh and dry weight of the fourth

experiment harvest. The ANOVA determined that the fresh head mass means were equal

throughout the system, and that the high or low density spacing pattern did not affect

growth. The ANOVA results of the fresh weight can be seen in figure 25. The p-value of

0.097 for the spacing treatment is above the benchmark of .05 and therefore the null

hypothesis of equal means cannot be rejected and the ability of plants to grow in either

special arrangement is confirmed. The fourth trial also confirmed that the face which

plants were grown on caused no difference in growth, with a p value of .064, indicating

no significant effect. The ANOVA for head dry weight is shown in figure 26, and

indicates the same as the fresh weight.

63
 Table 7. Total Head and Root Dry Mass, and Average Head and Root fresh Mass for 10 plants in
the fourth experiment harvest. Calculations were done in the same fashion as described in table
1

Fresh Weight W2 W3 W4 W5 W6 W8 W9 W10 E2 E3 E4 E5 E6 E8 E9 E10

Total Head
Mass / Row
(g) 1675 2165 1995 2125 1515 2040 2025 1695 2045 2205 1825 2135 2180 2300 2665 1625
Total Root
Mass (g) 115.0 300.0 275.0 300.0 180.0 415.0 365.0 235.0 325.0 405.0 300.0 245.0 600.0 300.0 330.0 280.0
Avg. Head
Mass (g) 167.5 216.5 199.5 212.5 151.5 204 202.5 169.5 204.5 220.5 182.5 213.5 218 230 266.5 162.5
Average Root
Mass per
Head (g) 11.50 30.00 27.50 30.00 18.00 21.84 19.21 12.37 17.11 21.32 15.79 12.89 31.58 30.00 33.00 28.00
Root Mass
Percentage 7% 14% 14% 14% 12% 11% 9% 7% 8% 10% 9% 6% 14% 13% 12% 17%

Table 8. Total Head and Root Dry Mass, and Average Head and Root Dry Mass for 10 plants in
the fourth experiment harvest. Calculations were done in the same fashion as described in table
1
Dry Weight W2 W3 W4 W5 W6 W8 W9 W10 E2 E3 E4 E5 E6 E8 E9 E10
Total Head
Mass / Row
(g) 48.77 56.42 50.36 53.88 50.21 58.04 59.71 54.23 53.28 53.06 51.9 53.87 58.09 60.59 61.36 50.87
Total Root
Mass (g) 11.2 22.2 17.6 17.9 14.0 25.2 23.1 17.9 19.5 22.7 19.8 14.5 21.4 18.2 20.6 18.3
Avg. Head
Mass (g) 16.26 18.81 16.79 17.96 16.74 19.35 19.9 18.08 17.76 17.69 17.3 17.96 19.36 20.2 20.45 16.957
Average Root
Mass per
Head (g) 1.12 2.22 1.76 1.79 1.40 1.33 1.22 0.94 1.03 1.19 1.04 0.77 1.13 1.82 2.06 1.83
Root Mass
Percentage 7% 12% 10% 10% 8% 7% 6% 5% 6% 7% 6% 4% 6% 9% 10% 11%

64
 350

300

250

200

150

100

50

0
W2 W3 W4 W5 W6 W8 W9 W10 E2 E3 E4 E5 E6 E8 E9 E10

Series1 Series2 Series3 Series4 Series5

Series6 Series7 Series8 Series9 Series10

Figure 24. Fresh lettuce head weight for 10 samples (1 – 10) from the fourth experiment
for each row number (2 through 10), and each face (East (E) or West (W)).The colors
within each grouping represents the different rows.

Figure 25. The ANOVA output of the SAS software comparing equal means among
spacing arrangements, fresh weight.

65
Figure 26. The ANOVA output of the SAS software comparing equal means among
spacing arrangements, dry weight.

The four experiments provided the expected production outputs under optimal

conditions. The operational inputs included the electrical consumption of the circulation

pump, which running continuously, was 5.5 kWh day-1, or 154 kWh for a 28-day

production period.

Nutrient solution consumption was approximately 4 liters per head of lettuce over

the total growth period, and was an average 3.6, 4.2, and 4.4 l min-1 for the second, third

and fourth experiments, respectively

The harvest rate of the lettuce heads were estimated to be 95% for each of the

second, third and fourth experiments.

Economic Model

The feasibility model for a hypothetical urban agriculture operation was created to

allow simple inputs to feed into pre-determined formulas which give the user a clear

66
output of costs and profit from the site. The basis for the model is the validated a-frame

hydroponic system discussed above which defines how the grow space inputted by the

user will be utilized and from this determination, provides an exact production model that

can be adapted to multiple cities and locations.

As the model was being developed, the largest constraint became clarity. To

preserve the simplicity of the ‘main’ sheet, auxiliary sheets were added to the model to

perform and explain to the user the calculations that ultimately feed into the cost analysis

for the operation. The auxiliary sheets serve two functions, the first is to provide

justification for all outputs on the main page and the second being to allow for greater

customization and less forced assumptions as can be seen in similar greenhouse pricing

models (Donnell, et al., 2011). The customization gives users experienced in the field the

ability to decide the exact inputs into the model on every level while still giving the less

experienced user the ability to use widely accepted assumptions in the industry to give an

accurate prediction of feasibility for a certain area. Nearly all cell formulas feed from

other cells, there are few unexplained numbers within the sheet that may skew financial

predictions when the user does not know to change them, the only exception being some

division by the number of months in a year and certain item pricing. The ‘assumptions of

the system’ portion was added to the main sheet to give the user more clarification on the

source of values used later in calculations. The assumptions include the correction

multipliers for over-seeding and spoilage prediction, with values of 1.05 and .90,

respectively. These multipliers create a conservative outlook of production and give more

credence to the production numbers being representative of a real-world facility.

67
 The most difficult input item for any potential user will be the heating cost

auxiliary sheet, considering the weather data required. Although this weather data is often

easily available to any user through federal and state-run weather reporting programs, it

could be streamlined for future use through real-time incorporation of a weather service

data system. Many of the other inputs are easily available to all consumers and require

minimal effort beyond entering the pricing or rate value into the ‘input’ section of the

main sheet. The ‘A-frame build cost’ and ‘nutrient solution calculator’ auxiliary sheets

are an accurate representation of current pricing but may require adjustment in the future.

A benefit of this feasibility model is its basis upon ‘cycles’ as the base unit of

time. Similar financial budget models often rely on standard dates, either quarterly,

monthly, or yearly, however, this time scale does not accurately represent the cash flow

of a system utilizing the a-frame hydroponic design. The cycle system is determined by

the user as the maximum number of rotations that an individual a-frame system can grow

a crop in a 365 day period. Due to varying weather and complexity of acquiring accurate

temperature and solar data for any given area, using a growth model would be far too

difficult for the average user. It would also be unnecessary for the advanced grower who

has the ability to determine number of crop cycles for their given area already.

The goal of the research is the output given in the cost breakdown. The use of

cycles as the time determinant allows the user to see their costs of operation both on a per

crop basis and a per year basis. The number of years observed is an input by the user

which determines total number of cycles for which the system will be in use before

replacement of the hydroponic a-frames. The total number of cycles, which is a

combination of cycles per year and number of years observed, feeds into the model to

68
determine the rate at which capital expenses such as greenhouse construction costs and

the hydroponic system construction costs will be paid off.

A ‘year 1 analysis’ was added to the main sheet of the calculator to provide the

user with a complete perspective on cash flow of the operation. The analysis delves into

all of the associated costs and returns of the operation, while proportioning each on a per

head, m2, system, and whole operation basis. This approach can be used to identify

strengths and weaknesses of a current pricing plan and give the user greater detail of their

operation.

Figure 27 is an example of inputs to the system of a standard commercial

greenhouse, from these inputs, the feasibility model then uses the calculations of

auxiliary sheets, formulas within the main sheet, and assumptions of the system to model

the situation. The facility designations and crop sections are shown in figure 28 with

outputs based upon the assumptions of the system which were determined from the crop

trials and empirical data. From these determinations, the cost breakdown is formed,

giving the user a view of both variable and fixed costs, displayed in figure 29. The model

then computes the revenues based upon maximum number of heads which may be grown

in the allotted space and the market price for a head of lettuce. The cost analysis is the

final output of the model, giving a value for return on investment for the current situation

while also outlining the source of costs and returns in the year 1 analysis, as shown in

figure 30.

A sensitivity analysis was added to the feasibility model to further the power of

the model to predict the financial outcome of a specific urban agriculture operation. The

sensitivity analysis functions through the interpolation of two chosen inputs, in this case,

69
growing space and the market price of a head of lettuce. Figure 31 provides the

sensitivity analysis from the situation described above and shows the price point, >$2.00

per head of lettuce, where no matter the size of the operation, any price below will not be

profitable. The price per head of lettuce is stated along the x axis while the square

meterage of the growing operation is listed along the y axis, this sensitivity table is

adaptable within the model document to allow the user to see a wide range of pricing and

size options and where, according to all other inputs held constant, they may expect to

produce a profit. The sensitivity analysis can be done on both a per crop and per year

basis with the ability to raise or lower the bounds of either parameter to any desired level.

Figure 27. The input area of the main sheet of the feasibility model. Sixteen input points

allow the user to form an accurate model of a hypothetical urban agriculture operation.

70
Figure 28. The facility designations area outlines the basic functions calculated from the
input area which feed later into the model. The crop section, both in cycle and per year,
give the user a clear readout of capabilities of the hypothetical facility.

Figure 29. Cost analysis calculations of the feasibility model under a hypothetical
situation models both variable and fixed costs of an operation.

71
 Revenues
Revenue per Cycle
Price Point ($) Revenue per Year ($)
150-200 (g) Head of Lettuce 15200 $ 2.50 ($/Head) $ 38,000.00 $ 228,000.00

Return on Investment Profit per Cycle ($) Profit per Year ($)
$5,937.54 $35,625.25

Year 1 Analysis
Return Over
Variable Costs Return on
Variable Costs ($) Fixed Costs($) Total Costs($) ($) Investment ($)
Per Head $ 0.71 $ 1.40 $ 2.11 $ 1.79 $ 0.39
Per m2 $ 322.18 $ 639.69 $ 961.87 $ 817.82 $ 178.13

Per System $ 2,577.46 $ 5,117.53 $ 7,694.99 $ 6,542.54 $ 1,425.01

Per
Operation $ 64,436.62 $ 127,938.13 $ 192,374.75 $ 163,563.38 $ 35,625.25

Figure 30. The revenues, profit and year 1 analysis portions of the feasibility model give
the user an area to adjust the market price of a head of lettuce to find an appropriate profit
margin. The year 1 analysis gives a breakdown of profit and cost areas within the
operation on a per unit basis.

Sensitivity Analysis (Price vs Growing space (m2))

$10,823.87 $ 1.00 $ 1.25 $ 1.50 $ 1.75 $ 2.00 $ 2.25 $ 2.50 $ 2.75 $ 3.00
100 $ (5,993.06) $ (4,093.06) $ (2,193.06) $ (293.06) $ 1,606.94 $ 3,506.94 $ 5,406.94 $ 7,306.94 $ 9,206.94
110 $ (6,591.37) $ (4,501.37) $ (2,411.37) $ (321.37) $ 1,768.63 $ 3,858.63 $ 5,948.63 $ 8,038.63 $ 10,128.63
120 $ (7,189.68) $ (4,909.68) $ (2,629.68) $ (349.68) $ 1,930.32 $ 4,210.32 $ 6,490.32 $ 8,770.32 $ 11,050.32
130 $ (7,787.98) $ (5,317.98) $ (2,847.98) $ (377.98) $ 2,092.02 $ 4,562.02 $ 7,032.02 $ 9,502.02 $ 11,972.02
140 $ (8,386.29) $ (5,726.29) $ (3,066.29) $ (406.29) $ 2,253.71 $ 4,913.71 $ 7,573.71 $ 10,233.71 $ 12,893.71
150 $ (8,984.59) $ (6,134.59) $ (3,284.59) $ (434.59) $ 2,415.41 $ 5,265.41 $ 8,115.41 $ 10,965.41 $ 13,815.41
160 $ (9,582.90) $ (6,542.90) $ (3,502.90) $ (462.90) $ 2,577.10 $ 5,617.10 $ 8,657.10 $ 11,697.10 $ 14,737.10
170 $ (10,181.21) $ (6,951.21) $ (3,721.21) $ (491.21) $ 2,738.79 $ 5,968.79 $ 9,198.79 $ 12,428.79 $ 15,658.79
180 $ (10,779.51) $ (7,359.51) $ (3,939.51) $ (519.51) $ 2,900.49 $ 6,320.49 $ 9,740.49 $ 13,160.49 $ 16,580.49
190 $ (11,377.82) $ (7,767.82) $ (4,157.82) $ (547.82) $ 3,062.18 $ 6,672.18 $ 10,282.18 $ 13,892.18 $ 17,502.18
200 $ (11,976.13) $ (8,176.13) $ (4,376.13) $ (576.13) $ 3,223.87 $ 7,023.87 $ 10,823.87 $ 14,623.87 $ 18,423.87

Figure 31. The sensitivity analysis of the feasibility model for a hypothetical urban
agriculture operation. Allows better perspective to the user for determining growing
space and price point as the two major influencers of profit.

Future research could be incorporated into the feasibility model to improve upon

certain features which are either based upon assumption or standard industry practices.

The heating cost calculator operates under the principle of user-inputted weather values

for their specific area for one year. At the current generation, the model assumes a

standard yearly climate for the total number of years observed. Future models could

incorporate real time weather inputs from governmental or research institution weather

72
collection services and could feed these values into the heating cost calculations. The

ability to compare different hydroponic system styles, such as deep water culture versus

nutrient film technique, would be another point of improvement for the model.

Comparison of other hydroponic systems is beyond the goals of this research, so it was

not included but it could be done through alteration of system assumptions and certain

cost factors. Investors may also be interested in partitioning crop sales between wholesale

and premium market pricing, however, this is beyond the scope of this research and

would involve major changes to the cost breakdown sector.

73
 CONCLUSION

The economic model is adaptable to a wide array of locations due to the flexibility

afforded by auxiliary sheets. The use of auxiliary sheets improved clarity and

customization for the user. A net profit was determined by use of the model for a market

price of $2.00 per head of lettuce when incorporating inputs from realistic scenario

greenhouses, and empirical and commercial data. . The economic model delivered a

preliminary view of an urban agriculture operation providing investors information about

the viability of a proposed business venture.

The hydroponic design performed well, and exceeded previous NFT A-frame

designs in the number of heads per system within a given footprint (Kacheris, 2014). The

low weight of the system from the small reservoir and aluminum structure were well

suited for a rooftop in addition to its effective use of three dimensional space for crop

production. The high density spacing of 79.5 heads m-2 provided a uniform and quality

harvest. The multiple trials performed with the system confirmed plant production

uniformity and the ability to use the system as a basis for the output area of the economic

model, and not having to use unreliable empirical data. This will provide a grower the

ability to utilize this system and have confidence in economic predictions.

74
References

Aldrich, R.A. and Bartok, J.W., 1994. Greenhouse engineering. NRAES (USA).

Bakker, N., Dubbeling, M., Gündel, S.A.B.I.N.E., Sabel Koschella, U. and Zeeuw, H.D.,

2000. Growing cities, growing food: urban agriculture on the policy agenda. A reader on

urban agriculture. DSE.

Barbosa, G.L., Gadelha, F.D.A., Kublik, N., Proctor, A., Reichelm, L., Weissinger, E.,

Wohlleb, G.M. and Halden, R.U., 2015. Comparison of Land, Water, and Energy

Requirements of Lettuce Grown Using Hydroponic vs. Conventional Agricultural

Methods. International journal of environmental research and public health, 12(6),

pp.6879-6891.

Barham, J., Tropp, D., Enterline, K., Farbman, J., Fisk, J. and Kiraly, S., 2012. Regional

food hub resource guide (No. 145227).

Brechner, M., Both, A.J. and Staff, C.E.A., Hydroponic Lettuce Handbook.Cornell

Controlled Environment Agriculture.

Chan, K.W., 2007. Misconceptions and complexities in the study of China's cities:

Definitions, statistics, and implications. Eurasian Geography and Economics, 48(4),

pp.383-412.

75
Christie, C. B., and M. A. Nichols. "Aeroponics-a production system and research

tool." South Pacific Soilless Culture Conference-SPSCC 648. 2003.

Donnell, M., Short, T., Moore, R. and Draper, C., 2011. Hydroponic Greenhouse Lettuce

Enterprise Budget. Columbus, OH.

Drechsel, P. and Kunze, D. eds., 2001. Waste composting for urban and peri-urban

agriculture: Closing the rural-urban nutrient cycle in sub-Saharan Africa. CABI.

Food Marketing Institute. 2011. U.S. Grocery Shopper Trends, Food Marketing Institute:

Arlington, VA.

Getter, K.L. and Rowe, D.B., 2006. The role of extensive green roofs in sustainable

development. HortScience, 41(5), pp.1276-1285.

Graves, C.J., 1983. The nutrient film technique. Horticultural Reviews, Volume 5, pp.1-

44.

Giacomelli, G.A., 2002, July. Considerations for energy management of greenhouse

heating and cooling. In Southern Greenhouse Vegetable Growers Association

Conference.

76
Gumble, J., Berghage, R. and Stearns, D., 2015. Production and Financial Analyses of the

Rotating Living Wall, an Urban Agricultural System. Journal of Environmental

Protection, 6(9), p.1029.

Hendrickson, M.K. and Porth, M., 2012. Urban Agriculture—Best Practices and

Possibilities. University of Missouri, pp.1-52.

Hoagland, D.R. and Arnon, D.I., 1950. The water-culture method for growing plants

without soil. Circular. California Agricultural Experiment Station, 347(2nd edit).

Jensen, M.H. and Collins, W.L., 1985. Hydroponic vegetable production. Horticultural

Reviews, Volume 7, pp.483-558.

Kacheris, W. 2014. Evaluation of Capacity: A Novel 3D Hydroponic System Design.

University of Florida.

King, R.P., 2010. Comparing the structure, size, and performance of local and

mainstream food supply chains (Vol. 99). DIANE Publishing.

Kozak, A.E., 2015. Envisioning a self-sustaining city: The practice and paradigm of

urban farming in Shanghai. University of Colorado at Denver.

Kujawski, R. and Ryan, D. 2014. Fertilizing Trees and Shrubs. UMass Extension,

Landscape, Nursery and Urban Forestry Program.

77
University of Guelph. University School of Rural Planning and Development and Leinan,

V.A., 2002. Sustainable Livelihoods and the Role of Urban Agriculture as a Policy and

Household Strategy in Cienfuegos, Cuba. University of Guelph.

Leisinger, K.M., Schmitt, C. and Pandya, L., 2002. Six Billion and Counting: Population

and Food Security in the 21st Century (Washington, DC: International Food Policy

Research Institute, 2002). I am indebted to Todd Benson, an old friend and staff member

at the International Food Policy Research Institute, for better understanding issues

related to food security in sub-Saharan Africa, pp.4-6.

Licamele, J.D., 2009. Biomass production and nutrient dynamics in an aquaponics

system.

Lichter, D.T. and Brown, D.L., 2011. Rural America in an urban society: Changing

spatial and social boundaries. Annual Review of Sociology, 37, pp.565-592.

Low, S.A., Adalja, A., Beaulieu, E., Key, N., Martinez, S., Melton, A., Perez, A.,

Ralston, K., Stewart, H., Suttles, S. and Vogel, S., 2015. Trends in US local and regional

food systems. US Department of Agriculture, Economic Research Service. Washington,

DC.

78
Luxhoj, J.T. and Giacomelli, G.A., 1990. Comparison of labour standards for a

greenhouse tomato production system: a case study. International Journal of Operations

& Production Management, 10(3), pp.38-49.

Magigi, W., 2013. Urbanization and Its Impacts to Food Systems and Environmental

Sustainability in Urban Space: Evidence from Urban Agriculture Livelihoods in Dar es

Salaam, Tanzania. Journal of Environmental Protection, 4(10), p.1137.

Mendes, W., Balmer, K., Kaethler, T. and Rhoads, A., 2008. Using land inventories to

plan for urban agriculture: experiences from Portland and Vancouver. Journal of the

American Planning Association, 74(4), pp.435-449.

Mougeot, L.J., 2000. Urban agriculture: definition, presence, potentials and

risks. Growing cities, growing food: Urban agriculture on the policy agenda, pp.1-42.

Neuner, K., Kelly, S. and Raja, S., 2011. Planning to Eat? Innovative Local Government

Plans and Policies to Build Healthy Food Systems in the United States.

Novo, M.G. and Murphy, C., 2000. Urban agriculture in the city of Havana: A popular

response to a crisis. Growing cities, growing food. Urban agriculture on the policy

agenda. (ZEL), pp.329-346.

Un-Habitat, 2008. State of the World's Cities 2008-2009: Harmonious Cities. Earthscan.

79
Sanyé‐Mengual, E., Cerón‐Palma, I., Oliver‐Solà, J., Montero, J.I. and Rieradevall, J.,

2013. Environmental analysis of the logistics of agricultural products from roof top

greenhouses in Mediterranean urban areas. Journal of the Science of Food and

Agriculture, 93(1), pp.100-109.

Sheikh, B. A. "Hydroponics: key to sustain agriculture in water stressed and urban

environment." (2006): 53-57.

Soares, H.R., Silva, Ê.F.D.F., Silva, G.F.D., Pedrosa, E.M., Rolim, M.M. and Santos,

A.N., 2015. Lettuce growth and water consumption in NFT hydroponic system using

brackish water. Revista Brasileira de Engenharia Agrícola e Ambiental, 19(7), pp.636-

642.

Kristina, H.T., 2001. Organic status and dietary role of organoponicos in Cienfuegos.

Yi-Zhang, C. and Zhangen, Z., 2000. Shanghai: trends towards specialised and capital-

intensive urban agriculture. Growing cities, growing food, urban agriculture on the

policy agenda, pp.467-76.

80
Appendix

Appendix i: Jensen Hydroponic lettuce formulation:

81
Appendix ii: A-Frame Construction Build List:

82
Appendix iii: Pump Sizing Calculation:

83