design report a water generator water - calvin · pdf filei executive summary the goal of this...
TRANSCRIPT
© 2012
Team 5: Water from Air, Calvin College
DESIGN REPORT
ATMOSPHERIC WATER GENERATOR
WATER FROM AIR:
TEAM 5
Ben Niewenhuis - EE
Chris Shepperly - ME
Ryan Van Beek - ME
Eric Van Kooten – ME
9 May 2012
i
Executive Summary The goal of this senior design project is to design and prototype an atmospheric water generator,
a device which produces drinkable water from humid air. Special emphasis is given to energy
efficiency and compatibility with renewable energy sources.
This project is the culmination of the engineering program at Calvin College. It is conducted
within the context of a two-semester course which covers all aspects of project development and
management. The following report explores the feasibility of the proposed design as well as
specifying components of the design.
After careful research and testing, Team 5 has concluded that wet desiccation is not a practical
process for atmospheric water generation. The prototype works and is capable of producing 0.72
liters of water per day with significant potential for improvement. However, one of the metrics
used to compare this unit to comparable units is the water per unit energy. Water From Air can
produce 72.1 mL of water per kW-hr; Ecoloblue, a leading competitor, can produce 1031 mL of
per kW-hr. Given this factor of 10 difference, Team 5 has concluded that this design is
impractical for atmospheric water generation.
ii
Table of Contents
Introduction ............................................................................................................................. 1 1.
1.1. Background ...................................................................................................................... 1
1.2. Team Members ................................................................................................................. 1
1.2.1. Ben Niewenhuis ........................................................................................................ 1
1.2.2. Chris Shepperly ......................................................................................................... 1
1.2.3. Ryan Van Beek ......................................................................................................... 1
1.2.4. Eric Van Kooten ....................................................................................................... 2
1.3. Problem Statement ........................................................................................................... 2
Constraints ............................................................................................................................... 2 2.
2.1. Requirements .................................................................................................................... 2
2.2. Design Objectives ............................................................................................................ 3
2.3. Deliverables ...................................................................................................................... 3
Design Norms .......................................................................................................................... 3 3.
3.1. Transparency .................................................................................................................... 4
3.2. Stewardship ...................................................................................................................... 4
3.3. Integrity ............................................................................................................................ 4
3.4. Trust ................................................................................................................................. 4
Design Evaluation.................................................................................................................... 4 4.
4.1. Dehumidification .............................................................................................................. 4
4.1.1. Refrigeration ............................................................................................................. 5
4.1.2. Pressure ..................................................................................................................... 5
4.1.3. Combination .............................................................................................................. 8
4.1.4. Wet Desiccation ........................................................................................................ 8
4.1.5. Decision .................................................................................................................. 11
4.2. Brine ............................................................................................................................... 12
4.2.1. Corrosion................................................................................................................. 12
4.2.2. Cost ......................................................................................................................... 12
4.2.3. Safety ...................................................................................................................... 13
iii
4.2.4. Decision .................................................................................................................. 14
System Architecture .............................................................................................................. 14 5.
5.1. System Structure ............................................................................................................ 14
5.1.1. Frame ...................................................................................................................... 14
5.1.1. Air Blower Shelf ..................................................................................................... 15
5.1.2. Upper and Lower Shelves ....................................................................................... 15
5.2. Process ............................................................................................................................ 15
5.2.1. Airflow Loop .......................................................................................................... 16
5.2.2. Brine Loop .............................................................................................................. 17
5.2.3. Condensation Loop ................................................................................................. 20
5.1. Electronics ...................................................................................................................... 21
5.1.1. Power ...................................................................................................................... 21
5.1.2. Sensor ...................................................................................................................... 23
5.1.3. Control .................................................................................................................... 26
Prototype Operation ............................................................................................................... 26 6.
6.1. Description of Operation ................................................................................................ 26
6.1.1. Batch Process .......................................................................................................... 26
6.1.2. Manual Switches ..................................................................................................... 27
6.2. Operating Instructions .................................................................................................... 27
6.2.1. Start-Up Routine ..................................................................................................... 27
6.2.2. Shutdown Routine ................................................................................................... 28
Project Expenses .................................................................................................................... 28 7.
Testing ................................................................................................................................... 30 8.
8.1. Absorption Rate.............................................................................................................. 30
8.2. Evaporation Rate ............................................................................................................ 31
8.3. Prototype ........................................................................................................................ 32
8.3.1. Envirotronics ........................................................................................................... 32
8.3.2. Steelcase Inc............................................................................................................ 34
8.3.3. Produced Water Quality .......................................................................................... 35
8.3.4. Power Requirements ............................................................................................... 35
8.4. Control System ............................................................................................................... 36
iv
8.5. Further Testing ............................................................................................................... 36
8.5.1. Envirotronics ........................................................................................................... 36
8.5.2. Steelcase Inc............................................................................................................ 37
Project Design Improvements ............................................................................................... 37 9.
9.1. Financial-Based Improvements ...................................................................................... 37
9.2. Time-Based Improvements ............................................................................................ 38
Conclusion ......................................................................................................................... 39 10.
Acknowledgements ............................................................................................................ 39 11.
Appendix ............................................................................................................................ 42 12.
12.1. Competitor Summary ................................................................................................. 42
12.2. Decision Matrix .......................................................................................................... 42
12.3. Experiment Setup ....................................................................................................... 43
12.4. Experiment Data ......................................................................................................... 44
12.5. EES Calculations for Pressure Dehumification .......................................................... 45
12.6. EES Calculations for Desiccation Model ................................................................... 46
12.7. Complete Project Expenses Table .............................................................................. 47
12.8. Capacitive Sensor Circuit Reference .......................................................................... 48
12.9. Temperature Sensor Circuit Reference ....................................................................... 50
12.10. Envirotronics Validation Certificate ........................................................................... 54
Table of Acronyms
PPFS Project Proposal Feasibility Study
AWG Atmospheric Water Generator
WHO World Health Organization
CaCl2 Calcium Chloride
WBS Work Breakdown Schedule
MDF Medium Density Fiberboard
CFM Cubic Feet per Minute
EES Engineering Equation Solver
Table of Figures
Figure 1: Dehumidification by Refrigeration Cycle ....................................................................... 5
Figure 2: Dehumidification by Pressurization ................................................................................ 6
Figure 3: Power Requirements for Pressure Dehumification. ........................................................ 8
v
Figure 4: Dehumidification by Desiccation .................................................................................... 9
Figure 5: Research-Based Model .................................................................................................. 10
Figure 6: Results of the FEA analysis for the system’s frame ...................................................... 15
Figure 7: Block diagram of wet desiccation process .................................................................... 16
Figure 8: Candidate Locations for Sensor Placement ................................................................... 23
Figure 9: Humidity Sensor Circuit ................................................................................................ 24
Figure 10: Temperature Sensor Circuit......................................................................................... 25
Figure 11: Sensor Circuit Board Layout ....................................................................................... 25
Figure 12: Brine solution mass gain over time ............................................................................. 31
Figure 13: Temperature Curve for the Upper Tank ...................................................................... 32
Figure 14: Validation Graph of Envirotronics' Testing ................................................................ 34
Figure 15: Brine Concentration during Steelcase Testing ............................................................ 35
Figure 16: Experimental Setup ..................................................................................................... 43
Figure 17: EES Display for Pressure Dehumidification ............................................................... 45
Figure 18: EES Display for Desiccation Model ........................................................................... 46
Figure 19: PSpice Parametric Capacitive Sweep of Sensor Circuit ............................................. 48
Figure 20: Humidity Sensor Capacitance Response ..................................................................... 49
Figure 21: Sensor Capacitance Calculation .................................................................................. 49
Figure 22: Frequency to Relative Humidity Reference Chart ...................................................... 50
Figure 23: PSpice Voltage Sweep Simulation of Temperature Circuit ........................................ 50
Figure 24: Output Voltage vs. Ambient Temperature .................................................................. 51
Figure 25: Temperature Probe Calibration Results ...................................................................... 52
Figure 26: Temperature Circuit Calibration Mathcad Sheet (Screenshot) ................................... 53
Table of Tables
Table 1: Energy Values from Compression Condensation Base Case ........................................... 7
Table 2: Reported Nominal Operating Conditions and Performance ........................................... 10
Table 3: EES Model Results for Representative Conditions ........................................................ 11
Table 4: Salt Cost Comparisons .................................................................................................... 13
Table 5: Brine Decision Summary ................................................................................................ 14
Table 6: Component-wise Power Requirements of Prototype ...................................................... 22
Table 7: DC Power Supply Option Comparison .......................................................................... 22
Table 8: Main System Expenses. .................................................................................................. 28
Table 9: Breakdown of System Loop Expenses. .......................................................................... 29
Table 10: Summary of Design Features for Prominent Competition ........................................... 36
Table 11: Competitor Summary.................................................................................................... 42
Table 12: Desiccant Decision Matrix ........................................................................................... 42
Table 13: Recorded Project Expenses. .......................................................................................... 47
Table 14: Temperature Probe Calibration Data ............................................................................ 51
1
Introduction 1.
1.1. Background
Calvin College is a Christian Liberal Arts institution located in Grand Rapids, MI. It is one of the
few Christian colleges in America that offers a full engineering major. Integrating Christian
values and the Liberal Arts into engineering, Calvin’s program has a strong reputation for
producing thoughtful and well-rounded engineers.
This senior design project is the capstone course for the Calvin Engineering program. Two
courses, Engineering 339 and 340, are coordinated around the project, covering various aspects
of project management and design.
1.2. Team Members
1.2.1. Ben Niewenhuis
Ben Niewenhuis hails from Battle Creek, Michigan, a product of St Phillip Catholic High School
and the Battle Creek Area Math and Science Center. Active as a student and an athlete, Ben was
particularly interested in the realms of math and science. He entered the Calvin College
Engineering program in the fall of 2008, and has chosen Electrical Engineering as his field of
study. He is an active leader of the Calvin Chess Club and has taken two summer research
positions; one at Calvin College concerning polystyrene recycling and the other at Carnegie
Mellon University on improvements in Monte Carlo analysis. He looks forward to continuing his
studies at the graduate level in Electrical Engineering at Carnegie Mellon University.
1.2.2. Chris Shepperly
From Battle Creek, Michigan, Chris Shepperly chose to attend Calvin and pursue an engineering
degree. At Calvin Chris chose to concentrate in mechanical engineering because of his interest in
thermodynamics and mechanics of materials. Apart from engineering Chris spent a lot of time as
a swimmer for Calvin and was the captain for two years. His current plans are to enter the
workforce as a Systems Engineer at Dematic Corp. in Grand Rapids.
1.2.3. Ryan Van Beek
Ryan Van Beek is from Lansing, Illinois, where he graduated from Illiana Christian High
School. In the fall of 2008, Ryan began attending Calvin College and upon entering into his
junior year of college, Ryan began his focused study of mechanical engineering primarily in the
fields of machine design, thermodynamics, and heat transfer. Ryan’s engineering internship
experiences have been in both the aerospace industry at Woodward and HVAC modification
industry with Trane. After graduation in December of 2012 with a B.S.E. mechanical
concentration with a business minor, Ryan plans to work as a sales engineer for Trane while
continuing to live in the Grand Rapids Area.
2
1.2.4. Eric Van Kooten
Eric Van Kooten is a native of the Grand Rapids Area. He was born and raised in Kentwood,
Michigan and attended South Christian High School where he fostered his love for athletics,
being outdoors, and engineering. After South Christian Eric choose Calvin College and in order
to pursue engineering. Eric will graduate from Calvin College in January 2013 with a B.S.E.
mechanical concentration and International designation with a business minor. After Calvin, Eric
will work at Gentex and plans to utilize his international experience in industry and has a goal of
continuing his education with a dream of obtaining his doctorate and someday becoming
Professor Van Kooten.
1.3. Problem Statement
The initial problem statement for Team 5 was posed by Michael Harris, Executive Director of
the Calvin College Enterprise Center. The original statement was as follows:
“Design and develop a prototype system for removing clean (potable) drinking water from air
using a wind turbine. Use a wind turbine to generate electricity; use electricity to cool air (or
increase pressure) resulting in condensation of water; capture water vapor from air that
condenses into water. Self-contained system (no external connections required). Basic human
need (especially valuable for remote areas).”
Upon adoption of this project, Team 5 adapted the problem statement to better reflect the
realities of atmospheric water generation. Atmospheric water generators already exist as products
on the market; thus there is a need for this design to differentiate itself, through some innovation
or better utility in order to justify the expenditure of time and money on this project.
Furthermore, atmospheric water generation is an energy intensive process. Existing
implementations aren’t capable of producing significant amounts of water at a decent price. In
light of these considerations, Team 5 formulated the following problem statement:
“To design and prototype a system for obtaining clean drinking water from air, focusing on
improvements in the energy requirement with the end goal of powering the device with
renewable energy.”
Constraints 2.
2.1. Requirements
When designing the atmospheric water generator Team 5 identified three requirements they
needed meet to ensure that the final project would effectively fulfill its intended purpose.
Potability of Water - Water produced by the design must conform to the World Health
Organization (WHO) drinking water quality standards.
Simplicity of Use - Design must be operable by persons of limited technical experience.
Safety - Design must not pose a hazard to users at any point during its normal operation.
3
Potability of Water
The first requirement is that the water produced is safe to drink. This requirement is vital because
engineers must be socially responsible and take precautions so that society is not harmed by any
products they design. The World Health Organization (WHO) spends significant time working
globally to keep people safe from disease and other illnesses. Team 5 worked hard to follow
these standards so the end users are not harmed by the device they made.
Simplicity of Use
The second requirement is that the design is simple for people with limited technical experience
to use. The interface is designed to have approximately the same complexity level as standard
household appliances such as stoves, dishwashers, and washing machines.
Safety
In addition to the first two requirements, Team 5 also designed the AWG so that it is not
hazardous to users at any point of its standard operation. This third requirement is similar to the
first in that it seeks to protect the end users and society in general; however, this requirement
focuses on ensuring safety across all aspects of operating the system.
2.2. Design Objectives
When starting the project Team 5 developed several goals or objectives that the design will meet.
Flexibility in Power Source - The design should be able to utilize a variety of power
sources, including (but not limited to) solar, wind, and the traditional power grid.
1 Liter of Water Production per Day - The design should produce at least one liter of
drinkable water per day.
Maximize Efficiency - The design should maximize the water produced per unit energy.
Minimize Cost - The design should minimize the cost per unit water production for both
capital cost and production cost.
2.3. Deliverables
The final prototype of the atmospheric water generator will be a device that accomplishes the
requirements Team 5 established at the onset of the project within the scope and budget set forth
by the Senior Design Class. Team 5 will additionally provide a final design report, this
document, detailing the design process and specifications for the final design.
Design Norms 3.
For this project, Team 5 kept several design norms in mind as they planned to create this wet
desiccation system. This involved taking an honest look at the goals of the project and thought
on how the final product should interact with the users and other people who would be affected
by it.
4
3.1. Transparency
Throughout the design process we worked to make our process understandable, consistent, and
reliable. Someone will be using this so we should make sure that users without a technical
background can understand the process and make use of the product in everyday life.
3.2. Stewardship
The second design norm considered was stewardship. This norm is important for the project
because in using the earth’s resources, we must remember to take care of the world God has
given us. Using economic, environmental, and human resources in a manner that clearly
demonstrates a high value on stewardship shows we care about the world we live in and the
people we share it with.
3.3. Integrity
The design norm of integrity involves looking at the harmony between form and function,
completeness, promotion of human values and relationships, and is pleasing and intuitive to use.
Developing this AWG so that it accomplishes and reflects integrity in design meshes well with
the other design norms of trust and transparency. Our goal is to accomplish all of these design
norms to effectively make a better product that takes into consideration the rights of others.
3.4. Trust
The final design norm important for this project is trust. We want people to feel comfortable
using our AWG so that they use it to its fullest potential and their maximum benefit. Our desire
is to make a good product that is dependable and reliable. Making an unreliable product would
drive up user costs for repair and maintenance, our goal is to have a trustworthy design,
something that our customer can rely on.
Design Evaluation 4.
The system can be divided into four primary aspects: dehumidification, brine, control, and
power, each of which is discussed in detail below.
4.1. Dehumidification
When approaching the problem of atmospheric water generation it is clear that the heart of the
system is dehumidification, which is the removal of water from a stream of air. In this
application we seek to capture this water and utilize it for drinking purposes. Three common
psychometric methods of dehumidification stood out during preliminary research; a temperature
drop below the dew point (refrigeration condensing), pressure condensing, or a combination of
the two. In addition to these three psychometric methods, the team came upon an alternative
chemistry-related method called wet desiccation.
5
4.1.1. Refrigeration
Traditional refrigeration cycle dehumidification remains the most prevalent method for
generating water from atmospheric humidity. This method circulates air over cooling coils
connected in a refrigeration cycle to bring the water in the air below its dew point. The dew point
of the water is dependent on the vapor pressure and humidity and tends to be a relatively low
temperature compared to the ambient conditions. To reach the dew point the air running through
the unit will have to be cooled a considerable amount.1 This process requires a constant energy
supply that is used as the maximum allowable energy demand for the system. This approach is
expressed in Figure 1 below:
Figure 1: Dehumidification by Refrigeration Cycle
There are several advantages to this approach. First, it is founded on decades of technical work
and innovation. Furthermore, it is a very direct approach and relatively simple to evaluate given
psychometric theory and the latent heat of condensation. A primary disadvantage to this
approach is the magnitude of the heat transfer needed to generate a significant quantity of water.
Virtually all commercial atmospheric water generators utilize this approach to dehumidification.
4.1.2. Pressure
It is possible to compress humid air so much that it will condense at the ambient temperature. As
pressure increases the dew point rises; thus, enough compression will force the dew point above
the ambient temperature resulting in spontaneous condensation; heat will transfer from the
pressurized humid air to the ambient air. Compressing air to extract water could potentially
require pressures up to five times the ambient pressure. This will require a very sturdy tank that
can handle high amounts of stress in its walls. This method has great potential for low energy
1 Cengel, Yunus. Engineering Thermodynamics: Heat Transfer. Calvin College ed. Vol. 1. N.p.: McGraw-Hill,
2011. 192-202. 2 vols. Print.
Evaporator
Water
Humid Air Dehumidified
Air
Condenser
Warm Exhaust
Air
Compressor
Refrigerant
Cycle
6
demands, especially if one was able to recapture some of the energy in the compressed air using
a turbine or piston. The energy efficiency of this design option has great promise but it is heavily
dependent on compressor and decompressor efficiency and humidity. Figure 2 below is a
representation of this approach.
Figure 2: Dehumidification by Pressurization
The primary advantage of pressure dehumidification is the low energy requirement; the only
unavoidable loss is the pressure applied to the water vapor. However, any inefficiency in the
compression/decompression cycle is amplified by the large volume of air processed per unit
water produced. Additionally, the rate of production when driven by natural convection cooling
to the atmosphere is too slow for significant production; some mechanism to speed up this heat
transfer needs to be implemented, increasing the energy cost. No existing atmospheric water
generators utilize this approach.
Pressure assisted condensation was the first idea investigated in this project. Team 5 saw
potential in the alternate method of condensing because of the theoretical energy savings.
Because the team saw such promise for this approach thorough energy calculations were used to
evaluate the feasibility of the system. The system is deemed feasible if the energy per unit of
water is less than the competition. Several assumptions were made for the base case feasibility
calculations:
The incoming air has a relative humidity of 1 kgwater /kgdry air (ω1 =1) and the outgoing air
has a relative humidity of 0 kgwater /kgdry air (ω2 =0). This is a simplification because all air
will not come in with a humidity of 1 nor will it leave with no water vapor left in the air,
so this is an ideal humidity removal situation.
The compressor has a constant volume that does not change. A standard home
compressor has a volume of approximately 25 liters, which was used as the constant
volume for the base case.
Decompressor
Humid Air
Ambient Air
Dehumidified
Air
Pressure
Vessel
Compressor
Water
Heat
7
The mass of the water vapor removed from the air does not decrease the pressure in the
system at all. This is not reality because when the water is condensed out and then it is
pumped out of the control volume there will be less mass within the control volume and
the gas will be less dense. Since we have a constant volume the pressure will decrease to
account for the loss of mass.
The efficiency of the compressor was assumed to be 80%. This number is based on
preliminary findings from the internet.
The efficiency of the turbine was assumed to be 80%. The turbine and compressor were
assumed to be the same for simplicity but in reality they will vary.
The density of water is assumed to be 1000 kg/m3
for all temperatures. This was assumed
because the density variance over the small range of temperatures analyzed in this system
was negligible. Likewise the density of air was assumed to be 1.2 kg/ m3
for all
temperatures, for the same reasons.
Compression would not cause any change in temperature and the system would only have
to remove the energy required to condense the water, the latent heat of condensation. It
was also assumed that the heat could be removed in a perfectly efficient process.
For these assumptions Team 5 was able to calculate the amount of energy that was required per
mass of water. This unit mass of energy was found using Equation 1:
(Equation 1)
In is the electrical energy required and is the mass of the water. The total energy
required, energy recovered, and the energy required from the grid are related in Equation 2.
(Equation 2)
With all of the assumptions, Team 5 found that for each gram of water produced the system
would require 20,670 Joules from the grid which is more than the competition. Table 1 shows the
values for the energy calculations of the pressure condensing system along with basic
comparisons of the competitor’s energy requirements.
Table 1: Energy Values from Compression Condensation Base Case
Total Energy Required 57416 kJ/kg
Energy Recovered 36747 kJ/kg
Energy Required from Grid 20669 kJ/kg
Ecoloblue 3500 kJ/kg
8
After the base case calculations were complete, Team 5 was able to vary several parameters to
determine the minimum efficiency required in the compressor and turbine to equal the best
competitor’s energy requirement. First by varying the efficiency of the compressor and turbine as
one component, an average component efficiency of 92% is required to match Ecoloblue. When
the final pressure is lowered the average efficiency decreases up until a certain point. This shows
that there is an ideal pressure at which to operate the system. This ideal pressure is 190 kPa and
requires a minimum average component efficiency of 85%, Figure 3 shows the decreasing power
requirements for increasing average component efficiencies. Calculations for the compression
condensing system can be found in section 12.5 of the appendix.
Figure 3: Power Requirements for Pressure Dehumification.
4.1.3. Combination
The third alternative is a combination of compression and cooling. When humid air is
pressurized the dew point of the water vapor is increased. This is beneficial for an atmospheric
water generator because pressurizing the air a small amount so that less cooling is required could
lower energy requirements. By combining the two approaches it minimizes the work needed for
each to achieve the same desired result; however this would significantly increase the system
cost and complexity.
4.1.4. Wet Desiccation
The final design option considered is the most abstract, but has gained recognition recently as a
valid design for atmospheric water generation. Wet desiccation is a process where a brine
solution is exposed to humid air in order to absorb water vapor from that air. The solution is then
sent into a regenerator where the water vapor is extracted from the solution.2 This method has
2 Ramachandran, Balakrishnan. 10 Eco-Friendly Gadgets that Procude Water from Air. N.p., 10 Aug. 2011. Web. 8
Dec. 2011. <http://www.ecofriend.com/entry/10-eco-friendly-gadgets-produce-water-air/>.
0.76 0.8 0.84 0.88 0.92 0.96 10
1000
2000
3000
4000
5000
6000
7000
8000
hcompressor
kJ
/kg
qgridqgrid
qcompetitorqcompetitor
9
grown in popularity because of its efficiency and the ease with which it can be adapted to
renewable energy, particularly solar. Figure 4 below is a basic representation of this approach.
Figure 4: Dehumidification by Desiccation
A primary advantage to this approach is that the desiccant accomplishes the most difficult part of
dehumidification, extracting the water from the air, without a direct expenditure of energy. The
problem is thus recast into terms of regenerating the desiccant and capturing the resultant water.
The main disadvantage of wet desiccation is the complexity that is introduced, both in terms of
system and materials.
Actual implementation of wet desiccation dehumidification depends largely upon the desiccant
used. Two atmospheric water generators that utilize desiccation have been identified, each of
which utilizes a different material as its desiccant. The first is in development at the Interfacial
Engineering and Biotechnology IGB in Stuttgart, Germany and uses a highly concentrated brine
solution. The solution adsorbs water from humid air and is then cycled into the interior of the
device, where the water is extracted under low pressure conditions3. The other device uses a
solid-state desiccant and a day-night cycle; by night humid air is circulated, by day solar heat is
used to extract the water and regenerate the desiccant4.
Wet desiccation originally appeared to be a daunting task; however, as more research was
performed; wet desiccation became more appealing as a design alternative due to its potential for
low energy operation and room for innovation. The biggest difficulty was establishing the
feasibility of this approach, especially concerning the rate at which the water was absorbed by
the desiccant. Research yielded several useful articles that described the performance and
implementation of wet desiccation systems. Of these, one article in particular was best suited to
our purposes. This article described a cross-flow liquid desiccant dehumidification system and
reported the performance of the system across a variety of experimental conditions. The
operation of the system under nominal conditions was highly encouraging, a state summarized in
Table 2.
3 Fraunhofer-Gesellschaft. "Drinking Water From Air Humidity." ScienceDaily, 5 Jun. 2009. Web. 8 Dec. 2011.
<http://www.sciencedaily.com/releases/2009/06/090605091856.htm> 4 Ellsworth, Joe. Solar Thermal Air to Water. N.p., 28 Aug. 2011. Web. 8 Dec. 2011. <http://a2wh.com/>.
Humid Air Dehumidified Air Dehumidifier
Regenerator
Desiccant
Cycle
Water
10
Table 2: Reported Nominal Operating Conditions and Performance5
Parameter Value
Solution Temperature 30C
Solution Concentration (mass %) 43%
Solution Flow Rate 0.175kg/s
Air Temperature 30C
Air Relative Humidity 78%
Air Flow Rate 204.4cfm
Water Absorption Rate 79.49kg/day
Several observations can be made from these reported results. First, the rate of water production,
almost 80kg of water per day, translates into a volume much greater than our stated goal of 1L
per day. This is particularly encouraging considering the second observation: the operating
points of the reported system are tailored to provide maximum dehumidification performance.
Thus, in adapting the results of this system to our design one can only expect the absorption rate
to fall as compromises are made to reduce energy cost and maximize the water produced in the
regenerator system; the initial figure of 80kg per day gives good reason to believe that the
original goal can still be achieved in a similar system.
The next step was to construct a model using the relationships reported in this article between the
various operating conditions and the water absorption rate. Four key factors were identified: the
size of the contact surface, the rate of air flow, the concentration of the brine exposed to the air
stream, and the temperature of the brine. Linear approximations were developed for each of these
factors using the reported data. Additionally, it was assumed that each of these factors has an
independent effect on the system. Figure 5 below is a representation of the resultant model.
Figure 5: Research-Based Model
This model takes the four input variables noted above and returns a predicted rate of water
absorption and a nominal brine flow rate (flow rates below this nominal value fall off due to
incomplete wetting of the packing tower). This model was implemented in Engineering Equation
5 C. Moon, P. Bansal, S. Jain, “New Mas Transfer Performance Data of a Cross-Flow Liquid Desiccant
Dehumidification System,” International Journal of Refrigeration 32, 2009, p. 524-533.
Packing Dimension l
Brine Concentration C
Brine Temperature T
Air Flow
Model Water Absorption
Brine Flow
11
Solver (see section 12.6) and used to evaluate an operating condition representative of our
system design. Table 3 below summarizes the results of this simulation.
Table 3: EES Model Results for Representative Conditions
Parameter Value
Solution Temperature 30 C
Solution Concentration (w%) 32%
Packing Dimension 30cm
Air Flow Rate 93.5cfm
Water Absorption Rate 18.06kg/day
Nominal Brine Flow Rate 0.171kg/s
The critical change between the operating conditions and those used in the article is the solution
concentration: we will be dehumidifying with a concentration much less than that used in the
article. In this calculation 32% was the estimate used. Additionally, airflow was reduced to save
energy and lower noise. According to this model, even with these changes, 18kg of water can be
absorbed per day. Given that components supplying the required flow rates and packing
dimensions are available within the budget range (see Table 13), it is concluded that this
approach is feasible from a research perspective.
4.1.5. Decision
After careful consideration, Team 5 chose the wet desiccation system as the best option for
atmospheric water generation. Teams 5’s initial inclination was to work with independent
compression, but after performing energy calculations and going through careful consideration of
the alternatives it no was no longer favored. A main difficulty with compression is that in order
to make it competitive, in terms of energy usage, a prohibitively expensive high efficiency
compressor is required. Because of the pricey compressor Team 5 faced a tradeoff between an
energy efficient system and a relatively inexpensive system with pure compression; therefore it
was passed up as the design of choice.
Another factor in the design decision is the fact that no patents or reports of an atmospheric
water generator that runs only on compression could be found. Even though we had calculations
to support that it is possible, the fact that no one uses that method made the team hesitant to
pursue the process. We decided against cooling coils because currently every main stream
atmospheric generator uses them. To achieve the goal of making this technology more available
and to prepare for the future of this market, the technology designed needs to be new and
possibly better than the current competitors. All of these factors are quantified in a decision
matrix located in the appendix in section 12.2.
12
4.2. Brine
In the desiccation process there are several different salts that will work in the brine solution.
Lithium bromide (LiBr), lithium chloride (LiCl), and calcium chloride (CaCl2) are all common
salts used in this process. To determine the best option, they were compared by several factors:
corrosion, cost, and safety. Each of the considerations is explained below.6
4.2.1. Corrosion
Salt solutions can cause accelerated corrosion on many different materials. Because it is
unavoidable for us to use a salt solution, corrosion must be minimized corrosion by choosing
corrosive resistant components and choosing a relatively non-corrosive salt. All three of the salts
that Team 5 is considering are fairly non-corrosive, but the nature of a salt solution will cause
accelerated corrosion compared to pure water.
To minimize corrosion, inhibitors can be added to the brine. Inhibitors are commonly used in
lithium solutions in which they are effective. No matter how effective the inhibitor is, if it
introduces toxic or harmful chemicals into the system Team 5 can’t use it (For more details refer
to the safety section); unfortunately, most inhibitors are harmful to humans. CaCl2 requires
inhibitors as frequently as the lithium-based brines, but uninhibited CaCl2 is more corrosive than
either LiBr or LiCL with inhibitors.
No matter which salt or which inhibitor is used, corrosion will occur with most metals. Because
of this as many plastic components will be used as possible.
4.2.2. Cost
Cost is important when selecting the brine because the system will require a significant amount
of brine in order to run. Cost prices for all three salts were found online and are reported in Table
4. Looking at the table it is clear that CaCl2 is the clear favorite. The design will require more
than one kilogram or salt so using either of the lithium salts would be a significant burden on
Team 5’s budget.
6 Studak, Joseph W., and John L. Peterson. "A Preliminary Evaluation of Alternative Liquid Desiccants for a Hybrid
Desiccant Air Conditioner." Austin, Texas: Center for Energy Studies, The University of Texas, 1988. N. pag. Print.
13
Table 4: Salt Cost Comparisons
Component Price Amount
Unit Price
($) (g) ($/kg)
LiCl7 19 100 190
LiBr8 59 100 590
CaCl29 23 22680 1.01
4.2.3. Safety
Team 5 is creating an AWG with the final goal of producing drinkable water. If this system is to
be effective the final output must be safe for humans to consume and safe to operate. The brine
solution has the potential to cause the design to be unsafe in both of these areas.
In this design the brine will be regenerated using evaporation; this means that no salt should
remain in the water vapor because it will not evaporate out of the brine. Even so, we do not want
to bring water into contact with a compound that is harmful to humans if ingested. The general
public would not understand that the design is safe and the resulting negative image would
hinder any hope of having a marketable product. Of the three salts considered above, CaCl2 is
the safest. Both lithium based compounds are safe to minor ingestion and exposure but repeated
exposure could cause harmful side effects. Because of this it is a great risk to use either LiCl or
LiBr.
The brine can cause the design to be unsafe during operation if the brine mist gets out of the
packed tower. Once in the air, it is hazardous if users inhale the mist or it comes into contact
with their eyes. All three of the salts considered will cause irritation to the eyes and lungs if they
exit the AWG apparatus. Since this is unavoidable with all three alternatives, Team 5 has
decided to minimize the chance of escaped brine mist by putting air filters over the air intake and
outlet of the packed tower. These filters have been chosen to minimize and eliminate the
possibility of escaping brine mist.10
11
12
7 "Lithium Chloride, 100gr." Science Company. N.p., 2011. Google. Web. 8 Dec. 2011.
<http://secure.sciencecompany.com/Lithium-Chloride-100g-P6357C670.aspx>. 8 "Acros Organics, Lithium Bromide, Anhydro 100gr." Neobits. N.p., 2009. Google. Web. 8 Dec. 2011.
<http://www.neobits.com/acros_organics>. 9 "Cargill Salt 50850 Calcium Chloride Pellets - 50lb." Ace Hardware. N.p., 2010. Google. Web. 8 Dec. 2011.
<http://www.acehardwareoutlet.com/>. 10
The Dow Chemical Company. "PELADOW* DG Calcium Chloride." Material Safety Data Sheet. N.p.: n.p.,
2009. N. pag. MSDS Solutions Center. Web. 7 Dec. 2011. <http://msds.com/>. 11
FMC Lithium. "Lithium Bromide Solution, Uninhibited." Material Safety Data Sheet. N.p.: n.p., 2010. N. pag.
MSDS Solutions Center. Web. 7 Dec. 2011. <http://msds.com/>. 12
Anachemla. "Lithium Chloride." Material Safety Data Sheet. N.p.: n.p., 2004. N. pag. MSDS Solutions Center.
Web. 7 Dec. 2011. <http://msds.com/>.
14
4.2.4. Decision
After considering the various factors for all three components, the chosen brine solution will be
composed of CaCl2. The considerations are shown in Table 5. Although it was not the best
option in all of the categories, CaCl2 is the safest and cheapest. If the design of the AWG is
unsafe to use then it is worthless; using harmful brine is not an option on this project. The design
must be cheap enough for customers to afford and with the amount of salt needed the lithium
salts our design price would inflate quickly.
Table 5: Brine Decision Summary
Consideration LiCl LiBr CaCl2
Corrosion Low Low Medium
Cost High High Low
Safety Poor Poor Good
System Architecture 5.
The wet desiccation system can be divided into three overall sections: system structure, process
and electronics.
5.1. System Structure
The system structure is the physical components that go into the system not including any
electrical components.
5.1.1. Frame
The entire system needed to be supported by some kind of frame; this frame was designed and
built to carry the load of all of the equipment for water generation. Calvin College’s Metal Shop
had an abundance of 1” hollow square tubing with 1/8” thickness walls. The entire frame was
fabricated with this material and strength tests were simulated using Autodesk Algor to assure
structural integrity. An image of the Algor test is shown Figure 6 where a uniform load of 40 lbf
was applied to the upper frame. The corresponding stress to the AISI 1005 steel was only 550
psi, which is well below its maximum strength. Therefore this frame is sufficient to support all
equipment necessary for the system and has a safety factor of roughly 70.
15
Figure 6: Results of the FEA analysis for the system’s frame
5.1.1. Air Blower Shelf
This shelf was added to the system to support the air blower, chimney, and condensation coils of
the system. This addition was made out of the same 1” hollow square tubing as the frame and
attached to the lower side of the frame outside of the primary supports (It is not shown in Figure
6).
5.1.2. Upper and Lower Shelves
The shelves were added to hold all of the equipment for the system between the supports of steel
frame. They were made out of medium density fiberboard (MDF) which was chosen because of
its availability and high strength.
5.2. Process
The process section is composed of the mechanical and chemical methods that generate water.
The process is divided into three loops: airflow, brine flow, and condenser. Figure 7 is a block
diagram showing all of the necessary parts for the system to help visualize the process portion of
the design.
16
Figure 7: Block diagram of wet desiccation process
5.2.1. Airflow Loop
The airflow loop begins with the blower pulling ambient air into the condenser chimney. This air
is channeled over a series of condenser coils and then blown through a diffuser into the packed
tower assembly. This is where the air comes into contact with brine from the brine flow loop.
The air then exits the packed tower assembly and returns to the environment. Stages in this loop
are denoted by the letter “A” in Figure 7 above.
5.2.1.1. Air Blower
The key stage for absorption comes as the brine is flowing down the packed tower while the
water vapor filled air is blown over it. Initial calculations determined the blower would need to
be capable of pushing at least 190 cfm (cubic feet per minute) into the packed tower in order to
meet the water production target. After extensive research to find the lowest cost unit that would
provide the necessary airflow, a Stanley blower fan was purchased that had three settings ranging
from 1280 cfm to 2180 cfm. Even though these flow rates are much higher than we need, the fan
17
will be a good fit because the large increase in the cross-sectional area of the air flow path will
lower the volumetric flow rate.
5.2.1.2. Inlet Fan Diffuser
The selected air blower had a limited distribution range which had to be increased to cover the
entire 32 inch by 12 inch face of the packed tower. A diffuser was designed in Autodesk Inventor
to fit the dimensional requirements of the air blower and the packed tower. The diffuser consists
of a shell built out of sheet metal with cardboard air dividers. The air dividers break the diffuser
shell into seven sections and each divider was made out of cardboard, covered in packaging tape
to protect the cardboard, and reinforced with duct tape to increase the rigidity of the dividers.
This simple design was chosen because the time needed to make and attach sheet metal dividers
was far too much. Cardboard is much less permanent, but this prototype is meant to be a proof-
of-concept and not a long term use unit. A production quality AWG would utilize a corrosion
resistant diffuser constructed with plastic and designed with higher strength and longer use in
mind.
5.2.1.3. Air Outlet Slats
With the fan blowing across the flow path of the brine loop, it is important to keep the brine
inside of the packed tower casing. The initial design had an air filter on the outside of the packed
tower case to capture any brine that would be blown out of the loop. This filter was insufficient
and was quickly destroyed when brine came into contact with it. The second iteration of the
design was to place fiberglass screens in front of the air outlet hole. This worked well but the
screen was blown out of the packed tower case when the air blower was turned on. This caused
the screen to be useless.
To contain all of the brine, outlet slats were created so that any droplets blown through would be
caught on these angled polycarbonate slats and allowed to fall back into the lower tank while the
air moved up and out of the tower casing. Minimizing the escape of brine from the tower casing
was also accomplished by creating a small square fence around the sprayer to direct brine flow
into the packed tower instead of towards the sides where it could flow down and out of the tank.
5.2.2. Brine Loop
Brine flow begins in the upper tank where it is released from the upper tank directly into the
lower tank. The brine pump then circulates the brine up to the spray nozzle at the top of the
packed tower assembly. Once the brine has gone through the packed tower several times and has
absorbed water from the air it is re-circulated to the upper tank; this is done with the brine pump
as well. There is an electronically controlled T-valve directing the brine flow between the spray
nozzle and the upper tank. Once in the upper tank again the brine is dried using two 100 watt
light bulbs as heating elements. Stages in this loop are denoted by the letter “B” in Figure 7
above.
18
5.2.2.1. Upper Tank
This portion of the AWG is for the evaporation of pure water from the brine solution. One of the
most important design considerations for this part was insulation. Since energy conservation is
one of the goals of this project, the energy put into this tank needs to be saved and reused
throughout operation. Adequate insulation also lowers the time needed to fully evaporate the
solution in the upper tank. This tank needs to be air tight, but accessible for maintenance so foam
weather stripping was attached to the top edges so that the top layer of the tank can be pressed
against it tightly. The top layer is anchored down by ratchet straps which were a great solution to
making the tank air tight, but still accessible.
The lid of the upper tank is where the brine inlet and outlet tubes, air inlet and outlet tubes and
both light sockets for the heating element enter the top tank. Because the top tank needs to be air
tight all of these holes need to be sealed. The top is made of acrylic with holes drilled into it.
5.2.2.2. Packed Tower
The design of this component greatly affects the efficiency of the entire AWG. Its function is to
maximize the surface area of contact between the brine and the air, this allows them to interact
more and increase the absorption. Therefore the more tightly the tower is packed, the greater the
contact surface area, the more humidity is absorbed out of the air flow through the system.
In addition to the packed tower, a shell was built to encase it fully so that no brine could leave
the system. This shell was made out of clear polycarbonate so the process will be more
transparent for education and observation of the system.
5.2.2.3. Lower Tank
This part of the design is only for short-term storage of the brine as it cycles through the packed
tower. To fulfill the two main functions of brine storage and packed tower support, the tank was
designed to hold a maximum of ten gallons and two supporting bridges were put into the lower
tank to hold the packed tower above the liquid’s maximum height. This tank and the packed
tower supports were all constructed out of polycarbonate and carefully sealed to make sure the
tank would not leak any of the stored brine onto the lower shelf.
5.2.2.4. Brine Pump
The brine pump is one of the main mechanical components of our system because it is used to
circulate our working fluid through the packed tower and up to the top tank. To meet the target
water output, the system model in EES (Engineering Equation Solver) required the pump be able
to pump at least two gallons per minute. Since the system frame was constructed with the upper
tank approximately five feet higher than the lower tank the pump also needed to be able to
maintain that flow rate while overcoming the head pressure. A third required characteristic of the
pump was that it be corrosion resistant. Product research and conversations with a pump
salesman determined that the March Magnetic Drive Pump was the solution for this application.
19
This pump is constructed to have only corrosion resistant materials (polypropylene, ceramic, and
Viton) in contact with the fluid. It also has a maximum head of thirteen feet and maximum flow
of five gallons per minute (gpm). This pump will provide more than four gpm of flow at five feet
of head based on the pump curve provided by the manufacturer. This unit more than fulfills the
specifications determined as sufficient for this system.
5.2.2.5. Brine Distributer
This portion of the AWG will play a pivotal role in the operation of the brine circulation. There
must be a minimum of 2 gallons per minute flowing over the packed tower at any time.
Initially a spray nozzle was chosen, it was an in-ground sprinkler head that has a continuous
stream in a circular area large enough to cover the packed tower. This was effective except for
leaving a large dry section in the middle of the packed tower and getting a lot of overspray on the
walls of the packed tower encasement. This was a problem because we did not have even
distribution and the overspray lead to brine getting out of the encasement.
To determine the optimal brine distribution over the packed tower, a grid of ice cube trays was
placed beneath the packed tower. The ice cube trays allowed us to determine where the brine was
falling and where it was missing. Using this we were able to determine the optimal brine
distribution method. The first method used to address the problem was to drape a towel over the
top of the packed tower. The spray nozzle would saturate the towel and the brine would then
flow down through the towel evenly flow down the walls of the packed tower. The towel worked
well at distributing the brine more consistently, there was however still a small dry spot in the
middle of the packed tower directly below the spray nozzle.
Adding the towel greatly increased the overspray that was landing on the packed tower
encasement and there was a need to address this. We decided to make a 3 inch high square that
would sit atop of the towel on the packed tower. This contained all of the spray from the spray
nozzle so that none of it reached the packed tower encasement.
After adding the towel to the top of the packed tower the flow was dangerously close to falling
below the minimum of 2 gallons per minute. By removing the spray nozzle completely and just
having the open tube letting brine flow openly onto the towel, which is now folded in half to
make it twice as thick, the flow rate is enough to consistently remain above this minimum flow
rate. Also with the thicker towel the flow does not go right through, a slight poop builds up on
top of the towel, contained by the square, until enough head builds up to equilibrate the flow rate
from the hose and the brine flowing down the packed tower. Because there is pool on top of the
packed tower we are ensuring even brine distribution. The towel also acts as a brine filter
collecting and large contaminates floating through it.
20
5.2.2.6. Heating Element
To evaporate the pure water out of the brine solution in the upper tank, a significant amount of
heat needs to be added. The final design was concluded as two 100 watt light bulbs. Testing has
gone into verifying that these bulbs will provide sufficient heat to evaporate the water (see
section 8.2 Evaporation Rate). They also keep the project cheap and help minimize energy
requirements. If this AWG were to be coupled with solar panels, these light bulbs as heating
elements could be entirely replaced by solar heat.
To fill this requirement, a solar water heater would be the best substitution of direct heat from a
200 Watt source. Solar water heaters generally are bought to contain up to 100 gallons in a loop,
but for this system only 5 gallons are needed. A small solar water heater could be used by
cycling the brine solution from the upper tank through the solar panels to heat the liquid easily to
50 degrees Celsius. This would require a new system design, but would be the best application
of renewable energy for this system.
5.2.2.7. System Tubing
Originally, the tubing meant for the entire system was 1/4 inch, but it was quickly realized that
this size restricts the flow of brine and brings the flow below two gallons per minute. The
solution was to implement 1/2 inch tubing to allow for greater flow. The 1/4 inch tubing is still
used, but only in the condensation loop.
5.2.2.8. T-Valve
The T-Valve selected for this system is a three way valve that can be toggled electronically. It
has two different flow settings; the default setting is when power supply is cut off. Brine is
always being pumped into the T-Valve by the brine pump; at the default position the brine leaves
the T-Valve and goes to the brine distributer to travel through the packed tower. When power is
applied to the T-Valve, the flow switches so that all of the brine is being pumped to the upper
tank. Most of the time the system will need to be pumping brine to the top of the packed tower
assembly and by making this orientation of the T-Valve the default, we are saving energy.
5.2.3. Condensation Loop
The condenser loop begins in the upper tank. The air within the tank is saturated with the water
that has been evaporated with the heating elements. This air is pumped out of the tank and
through condenser coils. In these coils heat is removed from the air causing the moisture to
condense out and into the water storage unit. The air is then pumped back into the tank and is
bubbled through the brine so that it can pull more water out of the brine. The condenser coils are
placed in an airflow chimney. The chimney is located at the inlet of the blower for the airflow
loop, this is done so that air will flow across the condenser coils and heat transfer will occur at an
accelerated rate due to forced convection. Stages in this loop are denoted by the letter “C” in
Figure 7 above.
21
5.2.3.1. Condenser Coils
The evaporated pure water that came out of the brine must be condensed so that it is useful and
can be captured. To do this effectively, twenty feet of copper tubing was purchased and shaped
into a helical coil. The copper tubing’s heat transfer properties are the reason it was chosen.
These coils are located inside the chimney and the pure water condensate is pumped through the
cycle using the air pump. This coil ends inside the water storage which doubles as a
condensation trap.
5.2.3.2. Air Pump
This pump was chosen because it is powerful enough to force the air from the upper tank through
the entire 20’ of copper tubing into the condensation trap and back to its source.
5.2.3.3. Chimney
The air chimney was added later in the design process, but it will improve the condensation cycle
greatly. Having the blower fan’s inlet air come through the chimney will increase the airflow
across the condenser coils and increase the heat rejection from the hot condenser air to the cooler
inlet air. This forced convection will improve the condensation rate and because it is close to the
frame and supported partly by the blower shelf, it provides a convenient way to support the
copper tubing.
5.2.3.4. Water Storage Condensation Trap
To capture the water that condenses inside condenser coils, the condensation trap has a vertical
drop from the coil inlet. This drop will let gravity take the condensed water to the bottom of the
storage tube and allow uncondensed water and air to return to the upper tank to repeat the cycle.
This water storage container is a clear PVC tube with an inlet and outlet on the top and a water
release valve on the bottom. Clear pipe was used for this component so that the final step of the
process can be visualized, and so that the water level can be monitored.
5.1. Electronics
The electronic section covers the power, sensor, and control systems necessary for the operation
of the design.
5.1.1. Power
The power systems are responsible for two primary functions. First, they must supply the
necessary voltages and currents to all of the design’s various components. Second, the power
systems need to be safe: electrocution is one of the greater hazards associated with this design,
and as such, safe power system design can significantly reduce this danger. Additionally, one of
the goals of this project is to produce a device that is compatible with renewable energy sources.
Unfortunately, it is outside of the scope and budget of this project to implement this design
consideration in our prototype. Table 6 below contains all of the power requirements for the
prototype.
22
Table 6: Component-wise Power Requirements of Prototype
Component Voltage (V) Current (A) AC/DC
Air Blower 120 0.9 AC
Brine Pump 120 1.0 AC
Air Pump 120 1.8 AC
Heating Element 120 2.4 AC
Brine Routing Valve 120 1.6 AC
Brine Return Valve 12 1.7 DC
Sensors 5 / ±12 0.2 DC
Given the component requirements listed in the table above, the prototype’s power systems will
need to provide one AC voltage and three distinct DC voltages. The AC voltage is easily
accessible from the grid. The DC voltages, however, require the inclusion of a DC power supply.
Two options were explored for this power supply: a standard computer power supply unit and a
board designed and assembled for this specific application. Table 7 below lists the estimated
costs and labor hours required for each option.
Table 7: DC Power Supply Option Comparison
Power Supply Option Estimated Cost Estimated Labor (Hours)
Computer $25 2
Custom $15 20
Not only was the computer power supply significantly more time effective without a prohibitive
increase in cost, the computer power supply was of better quality than any realistic custom
design, came in a self-contained unit, and was guaranteed to provide the required voltages and
currents. In light of these considerations, a computer power supply was purchased and used in
the prototype.
The other primary consideration in the power system was the safety of the user. Given the
voltages supplied to the prototype and the fact that the prototype would be filled with a highly
conductive fluid, the threat of shock to the user is not insignificant. In light of this, several design
decisions were made to minimize this risk. First, all power was routed through the air blower,
which had a built-in 10A breaker. Thus, in the event of a short, the breaker would throw and
prevent electrical fires. Second, the entire frame was tied to ground to prevent the possibility of
the frame going live and shocking any user who came into contact with it. Finally, a GFCI
(Ground Fault Current Interrupter) was incorporated in the power supply lines upstream of the
prototype. This device interrupts the current if any imbalance is detected between the in and out
power lines (i.e. if any current is leaking to the ground). All of these features taken together
ensure that the user is in no way at risk of dangerous electrical shock during the normal operation
of the prototype.
23
5.1.2. Sensor
Sensor systems allow the user and/or controller of the prototype to determine the operational
conditions of the process. Sensor systems are particularly vital for determining and keeping
within optimal control patterns. Looking at the prototype system diagram, there are six locations
where sensors would provide useful data, as shown below in Figure 8.
Figure 8: Candidate Locations for Sensor Placement
Every significant aspect of the system can be characterized by sensors placed at each of these
locations. The difference between the temperature and humidity readings at (I) and (II) would
provide the rate of absorption of water into the brine. A similar difference in readings between
(V) and (VI) would give the rate of water production. Brine temperature and conductivity
readings at (III) and (IV) would provide the basis for determining when to cycle the brine
solutions.
Limited resources, however, prevented the inclusion of sensors at every locations noted above in
the prototype. In particular, accurate conductivity sensors were prohibitively expensive ($67.95
II
I
III
IV
V
VI
24
for Hanna Instruments HI76128513
). In light of these restrictions, sensor placement was
prioritized. Temperature probes were placed in three locations (I, III, and IV above), and a
humidity sensor was placed at (I). These four measurements are sufficient for control of the
system once a control scheme has been determined, and establish a basis for inclusion of further
sensors if needed.
Components were chosen to meet these sensor requirements. Two specific components were
selected based on accuracy and cost: the temperature sensor integrated circuit MCP9701-E and
the capacitive humidity sensor HCH-1000-002, by Honeywell. A circuit was designed to accept
the output from each of these sensors and convey it to the user. In the case of the humidity sensor
such a circuit was necessary: the capacitance was on the order of 300 to 500 pF, meaning that a
highly accurate capacitance meter would otherwise be required. Figure 9 below is a PSpice
schematic of the circuit designed for this purpose.
Figure 9: Humidity Sensor Circuit
This circuit uses a 555 timer to translate minute differences in sensor capacitance into
measureable differences in output frequency. Refer to Appendix p. 48 for the relevant pages of
the component datasheet, PSpice simulation results, calculations, and the final reference chart for
the humidity sensor system.
The temperature probes, alternatively, did not require any additional circuitry; the voltage output
from the selected component was designed to be linearly dependent on the temperature.
However, because sensor output would be read by the user in the prototype, a circuit was
designed to scale and adjust the output voltage such that the conversion from voltage to
13
http://www.neobits.com/hanna_instruments_hi761285_conductivity_probe_with_built_in_temperature_sensor_for_
hi8730_p1829307.html?atc=gbs
0
X1
555D
GN
D1
TRIGGER2
OUTPUT3
RESET4
CONTROL5
THRESHOLD6
DISCHARGE7
VC
C8
V1
5Vdc
RA
100k
RB
100k
C_Sensor
{CSense}
C2
1u
R3
1k
PARAMETERS:
CSense = 300p
25
temperature could be performed without a reference chart. Figure 10 below is a PSpice
schematic of the circuit designed for this purpose.
Figure 10: Temperature Sensor Circuit
This circuit utilizes a follower circuit succeeded by an inverting amplifier, implemented using
operational amplifiers, to apply a gain and offset to the temperature IC output. Refer to Appendix
p. 50 for relevant pages of the component datasheet, PSpice simulation results, calculations, and
calibration data for the temperature sensor systems.
Three copies of the temperature sensor circuit and one copy of the humidity sensor circuit were
printed onto a custom circuit board, with layout shown below in Figure 11.
Figure 11: Sensor Circuit Board Layout
0
U2
uA741
+3
-2
V+7
V-
4
OUT6
OS11
OS25
V1
12Vdc
V2
12Vdc
VCC
VEE
VCC
VEEV5
0
V4
0Vdc
R1
799
R2
4201
R3
4664
R4
336
U3
uA741
+3
-2
V+7
V-
4
OUT6
OS11
OS25
VCC
VEE
V3
5Vdc
V5
Temperature Sensor
Potentiometer
Potentiometer
26
The above board was fabricated by members of Team 5 at Calvin’s facilities. This circuit board
and the corresponding sensors were successfully integrated into the final prototype.
5.1.3. Control
Control systems are responsible for making decisions as to the operation of the prototype. Team
5 has opted not to implement full electronic control in the prototype. Instead, the user will be
responsible for engaging the various valves and pumps in the prototype to achieve operating
conditions. This decision was motivated by several factors. First, electronic control is an
additional layer of complexity; user-based control simplifies the prototype electronics
considerably. Second, the exact nature of the required electronics is dependent on the desired
control. As Team 5 had no experience with controlling a wet desiccation process, the
requirements were unknown prior to prototype construction. Finally, Team 5 was limited
resources in terms of expertise (given only one electrical engineer on the team) and budget.
Electronic control would, however, be required in future versions of this system, particularly to
meet the Simplicity of Use requirement. Because of this, Team 5 designed the prototype with
future control improvements in mind. The current design brings together many of the sensors and
switches at a single location (the electronics faceplate) in the prototype. This would be the
location for any electronic controller in future versions. Manipulation of the various pumps and
valves would be achieved via relays controlled by a single microcontroller. This microcontroller
would have all of the sensors as inputs, along with a single on/off power switch from the user.
The microcontroller would thus be able to monitor and control the process via the sensors and
relays according to its programming, which would be determined based on optimizations of the
process. Finally, an LCD screen could be implemented in order to make the sensor data and
operating state of the system available to the user, increasing the transparency of the design.
Prototype Operation 6.
The prototype is constructed to be a proof-of-concept and it requires certain considerations while
it is operating.
6.1. Description of Operation
The following section of the report describes the prototype operation.
6.1.1. Batch Process
The brine loop operates in a batch process which requires the condenser loop to operate in a
batch process as well. This was chosen because space constraints on the packed tower meant the
brine would not absorb enough water in one cycle to justify circulating it to the upper tank.
Therefore the brine needs to be circulated for up to two hours through the packed tower assembly
until it is saturated enough to be sent to the upper tank for drying.
27
The actual times needed for each cycle were determined by testing the prototype and finding an
optimal use time. The brine loop requires a total of 2.5 hours to fully saturate. This cycle time
was determined at test conditions so actual cycle time would most likely be longer than 2.5
hours. Entirely independent of environmental conditions, the condensation loop is most effective
if run for 13.5 hours. This cycle much slower relative to the brine loop and requires an ample
amount of time to heat the brine solution and condense the water vapor back into liquid.
6.1.2. Manual Switches
Since the brine loop is operating as a batch process, switches must be flipped by hand at the
correct time to allow brine to stay in the packed tower cycle or pump to the upper tank. This
process could be automated and electronically controlled after extensive testing, but automation
is outside the scope of this project.
With time not available this semester, the prototype could be run through many different
operating conditions and flow rates to determine the time it would take to saturate brine. With
this optimization an electronic clock could be set to control the valves throughout system
operation. This clock or a concentration sensor would need to adapt to changing operating
conditions during operation. Once the concentration of the brine in the tank was diluted to a
specific level, the switch could flip automatically and send brine to the upper tank.
6.2. Operating Instructions
The following information is a step by step process of turning on and turning off the AWG.
6.2.1. Start-Up Routine
All components must start in an off position or unplugged, the following is a list of all switches
and plugs in the appropriate position.
Brine Return (Missile Switch) OFF (flipped up, cover up)
Brine Routing (Square Toggle) UP Position
Air Pump (Pin Toggle) UP Position
Power Supply OFF
Power Strip OFF
Lights Unplugged
Brine Pump Unplugged
The next list shows the order of turning on all systems assuming that the system has been set
according to the section above.
1. Plug in Air Blower
2. Turn on Power Strip
3. Turn on Power Supply
4. Plug in Brine Pump
28
5. Plug in Lights
6.2.2. Shutdown Routine
The following set of instructions shows the step by step instructions for shutting down the
system.
1. Unplug Brine Pump
2. Unplug Lights
3. Turn all Routing Switches to their OFF position (see above)
4. Turn OFF Power Supply
5. Turn OFF Power Strip
6. Unplug Air Blower
Once these steps have been accomplished, the entire system has been shut down and is safe to be
maintained.
Project Expenses 7.
In order to determine the overall expense of building an atmospheric water generator, the team
developed an original budget for the prototype by researching what components were available
both online and in the Calvin College metal shop. This budget served as a baseline to guide part
ordering and prototype construction. Primary component sizes were determined based on the
necessary performance levels as determined by the system model developed in EES (Engineering
Equation Solver). The brine pump, air blower, and packed tower were the three main
components and also the three highest individual expenses. More detailed descriptions of these
parts and others are included in the System Architecture section..
Table 8: Main System Expenses.
Main Systems
System Cost
Brine Loop $342.89
Air Loop $54.92
Condenser Loop $32.80
Electrical Systems $81.72
Other $0.00
Total Cost $512.33
The final overall expenses for this system are shown in Table 8. The original budget for Team 5
was five hundred dollars and after purchasing all of the parts for the project, the final expenses
exceeded the original budget by twelve dollars and thirty-three cents.
In Table 8, several main systems are mentioned; the more detailed system costs are shown in
Table 9. The sizes, manufacturers, and quantities of each component are specified in the
Appendix in Table 13. Donated components are listed as costing $0.00 in all expenses tables.
29
Table 9: Breakdown of System Loop Expenses.
Brine Loop
Components Cost
Magnetic Drive Pump $159.18
Packed Tower $67.10
Packed Tower Case $0.00
Lower Tank $0.00
Upper Tank $11.12
Insulation $0.00
CaCl2 $0.00
Spray Nozzles $3.60
Funnels $4.48
Fittings $35.57
System Tubing $25.89
Miscellaneous $18.27
Upper Tanks Straps $17.68
Total System Cost $342.89
Air Loop
Blower Fan $53.95
Air Filter $0.97
Diffuser $0.00
Total System Cost $54.92
Condenser Loop
Air Blower Pump $0.00
Copper Coils $18.14
Air Column Duct $0.00
Collector Tube Parts $5.19
Light Bulbs $9.47
Total System Cost $32.80
Electrical System
Humidity and Temperature Sensors $33.01
GFI Outlet $16.94
Switches and Power Supply $31.77
Total System Cost $81.72
Other
Metal Frame $0.00
MDF Shelves $0.00
Total System Cost $0.00
30
Testing 8.
Testing and experimentation allowed Team 5 to understand the relationship between different
process variables and how the system reacts to environmental changes.
8.1. Absorption Rate
In order to use wet desiccation it is essential to know the rate at which a brine solution can
absorb water. Team 5 met with Mr. Glenn Remelts and Professor Vander Griend to answer this
problem. Mr. Remelts helped lead the search for scholarly articles regarding brine absorption and
experimental setups. Professor Vander Griend helped the team develop the set up for several
experiments that will help to determine the absorption rate of a concentrated brine solution.
In the first experiment executed, a small mouth mug and a wide mouth mug (40% concentration
of CaCl2 by mass) were placed on a plate with water on it and covered with a larger pot. With
both mugs under the dome the atmosphere would remain unchanged except when measurements
were made. The water in the plate would evaporate into humidity which would then be absorbed
by the brine in the mug. A picture of the experiment is included in Appendix 12.3. The water
was held in a plate instead of a mug so that the test would not be limited by the speed at which
the water came out of the mug but by the rate at which it could get into the brine. This test took
place from December 1, 2011 to December 5, 2011 with the two different sized mugs so two
different absorption rates could be obtained based on each different mug. The data collected is
included in Appendix 12.4. The contact surface area between the brine solution and the
controlled atmosphere for the small mouth mug was 0.00422 m2 and 0.00633 m
2 for the wide
mouth mug. For each mug a function was developed for the change in mass over time. Looking
at Figure 12 shown below this relationship was different for the different surface areas. The
wide mouth mug has an absorption rate of 1.92 grams of water per day and the small mouth mug
has a rate of 1.00 grams of water per day.
31
Figure 12: Brine solution mass gain over time
Because there were two different surface areas, each mug gained different amounts of water over
the time of measurement, and a mass flux rate was calculated to compare the two more
accurately. The wide mouth mug had a flux rate of 303.2 g/day-m2 and the small mouth mug’s
flux rate was smaller at a rate of 235.8 g/day-m2. This comparison clearly shows that increasing
the surface area increases the rate which water is absorbed. Through research a company was
found who made a similar device with a 16.4 m2 absorption surface made of a cross corrugated
cellulose pad.
8.2. Evaporation Rate
The upper tank is designed to use heat sources to evaporate the excess water from the brine so
the vapor can be condensed down to liquid water. It is important to test this rate under operating
conditions to determine the effectiveness of the design.
Figure 13 shows the temperature curve of the upper tank during a test that was run on the
condenser loop with pure water in the tank. Temperature probes were placed in the upper tank to
record the change in temperature over time. The light heat source was turned on and the water in
the upper tank was allowed to heat up to 45 °C, this took roughly 150 minutes. At this point the
∆mass = 1.00t R² = 0.9998
∆mass = 1.92t R² = 0.9987
0
1
2
3
4
5
6
7
8
9
0:00:00 12:00:00 24:00:00 36:00:00 48:00:00 60:00:00 72:00:00 84:00:00 96:00:00 108:00:00
Mas
s G
ain
(g)
Time Elapsed, t (days)
CaCl2 Brine Solution Absorption Rate
Small Mouth Mug (Area = 0.00422 m^2)
Wide Mouth Mug (Area = 0.00633 m^2)
32
air pump was turned on and the air was bubbled through the water in the tank. As seen in Figure
13, this caused the temperature to jump nearly 5 °C, this leads us to believe that the temperature
probe was not taking accurate measurements and that the water needed to be stirred. The test was
run for another hour to condense water out of the air in the tank. In one hour of having the air
pump running 59.8 grams of water were produced. This gave us a baseline to work from.
Because this test was run with pure water we can expect the water production rate to be slightly
decreased but not significantly.
Figure 13: Temperature Curve for the Upper Tank
8.3. Prototype
The goal of our project is to generate water using the air surrounding the device. Temperature
and humidity are key variables that influence the rate of water production and testing them in a
measureable way is essential.
8.3.1. Envirotronics
Through a string of contacts beginning with Professor Harris and Steve Beukema, we were
connected with Dwayne Botruff who is the engineering manager at Envirotronics. Envirotronics
33
is an industry leader in the design, manufacture, and service of environmental test chambers
based in Grand Rapids14
. Dwayne provided Team 5 with the opportunity to use a test chamber to
manage the humidity and temperature during testing. Using this room for testing allowed the
team to subject the prototype to varying temperatures and humidity cyclically to simulate day
and night conditions. Testing the AWG in this room enabled the team to gather a much wider
range of performance data to develop a model for predicting system performance given various
atmospheric conditions. The result will be a greater ability to analyze the prototype and optimize
it to achieve maximum water production at a minimum cost.
In the visit that Team 5 made to Envirotronics, the only test conducted was on the brine loop.
The environmental test chamber that was used for testing was just larger than the size of the
prototype and it was impractical for the condenser loop to be run because of these size
limitations. The brine loop was run for six hours and every two hours three 50 mL samples of the
brine were collected. This was done in order to determine the concentration change of the brine
over the time of testing. These samples were weighed before testing and then were placed in an
oven to evaporate all of the water. Once this process is complete the mass of the sample, which is
now just salt, is measured again and this determines the change in concentration. Three samples
were taken each time to get an effective average.
The samples that were taken at Envirotronics were placed in the oven with a temperature that
was too high. All of the samples boiled over and were ruined, none of the testing from
Envirotronics can be measured but the testing was still valuable, Team 5 was able to see the
brine loop work for an extended period of time.
The final piece of documentation that Envirotronics left with us is validation certificates. Figure
14 shows the humidity and temperature inside the test chamber during the time of our testing.
This graph and the supporting validation certification (found in 12.10 Envirotronics Validation
Certificate) show that our tests were performed under the conditions specified and are accurate.
The sharp downward spikes in humidity correspond with us opening the door of the chamber
take data and samples, these humidity spikes correspond with a spike in temperature that is less
dramatic but still significant.
14
"About Us: Overview." Envirotronics: Environmental Test Chambers & Services. N.p., 2012. Web. 20 Apr. 2012.
<http://envirotronics.com>.
34
Figure 14: Validation Graph of Envirotronics' Testing
8.3.2. Steelcase Inc.
Through several more contacts Team 5 was connected with Marty Bender, manager of the
product development test labs at Steelcase Incorporated. Steelcase offered the use of their
humidity controlled chamber, a large room that is maintained at 85% relative humidity and 80
degrees Fahrenheit. With this chamber the entire prototype could be tested contrasting the tests
conducted at Envirotronics where only absorption was possible.
Like at Envirotronics, Team 5 took periodic brine samples in order to determine the brine
concentration in the lower tank. These were placed in the oven at a much lower temperature and
were correctly processed in order to determine the change in concentration. These samples were
dried for two days and then weighed. The results are graphed in Figure 15. This graph shows that
the brine becomes more concentrated as the tests move forward. This goes against what the
graph should show and what actually happens in reality. There are several possibilities as to why
this data seems to be incorrect. One idea is that the brine samples were not completely dry when
they were weighed and there is still some water trapped in the salt causing the sample to weigh
more and skew the results. Another possibility is that brine which had been dried in the upper
0
10
20
30
40
50
60
70
80
90
100
9:2
6:2
2 A
M
9:4
1:2
2 A
M
9:5
6:2
2 A
M
10
:11
:22 A
M
10
:26
:22 A
M
10
:41
:22 A
M
10
:56
:22 A
M
11
:11
:22 A
M
11
:26
:22 A
M
11
:41
:22 A
M
11
:56
:22 A
M
12
:11
:22 P
M
12
:26
:22 P
M
12
:41
:22 P
M
12
:56
:22 P
M
1:1
1:2
2 P
M
1:2
6:2
2 P
M
1:4
1:2
2 P
M
1:5
6:2
2 P
M
2:1
1:2
2 P
M
2:2
6:2
2 P
M
2:4
1:2
2 P
M
2:5
6:2
2 P
M
3:1
1:2
2 P
M
3:2
6:2
2 P
M
deg
."c"
& %
rh
date & time
Calvin College
tempsetpoint
actualtemp.
humiditysetpoint
actual%rh
35
tank was released to the lower tank. This brine is at a different concentration and will change the
overall the concentration of the lower tank.
Figure 15: Brine Concentration during Steelcase Testing
During testing the temperature in the upper tank and lower tank as well as the relative humidity
in the test chamber were taken every half hour. Using these measurements a time cycle was
determined.
The best cycle was run at Steelcase where 45 mL of water was produced in roughly 1.5 hours.
This was a significant production amount for Team 5 and it helped set some base values for
comparison of further improvements.
8.3.3. Produced Water Quality
Using evaporation and condensing processes will produce pure water by definition. Team 5 will
test the water produced to make sure there is no remaining Calcium Chloride in it. This is the
main concern with the drinking water. A secondary concern is the taste, smell and appearance of
the water.
8.3.4. Power Requirements
It is important to measure the power requirements of the prototype so that Team 5 can evaluate
the energy performance. One of the objectives of the project is to maximize efficiency, and Team
5 will measure this efficiency in units of water produced per unit energy. The basis for this
comparison is competitor models as described in Table 10. The energy use will be measured
using a watt meter. The watt meter measures the electrical power supplied to the system. This
watt meter can be used to measure the energy of the overall system but also the energy of several
of the components individually.
0.26
0.265
0.27
0.275
0.28
0.285
0.29
0.295
9:30 10:18 11:01 11:30 12:30
Bri
ne
Co
nce
ntr
atio
n
Time
Brine Concentration during Steelcase Testing
36
Table 10: Summary of Design Features for Prominent Competition
Device Manufacturer Production
(L/day) Power
(W)
Water Filtration
Stages
Unit Cost
Ecoloblue 2615 Ecoloblue 26 1050 11 $1,199
Dolphin116 Air2Water 22.8 500 5 $1,799
Atmos 2817 Atmos H2O 28 500 5 $1,545
8.4. Control System
One of the critical areas of testing will be determining what manner of control the system
requires for optimum operation. Mass production plans are to implement a series of control
statements using simplistic analog circuitry. The first order of business will be to determine the
thresholds required for good control on the system. From there, the most effective means of
applying the control signals will be determined, beginning with an assessment of the delays
inherent in the system process. Once these issues have been addressed, the overall performance
of this type of control will be examined, and further improvements added as necessary.
8.5. Further Testing
Team 5 wishes to perform further testing on the prototype and possibly revised prototypes. Some
of these tests are modifications to the tests that were performed and others are new tests that
were determined as valuable after the prototype was created.
8.5.1. Envirotronics
Testing improvements that Team 5 would like to make to the testing that was performed at
Envirotronics include a concentration sensor and humidity sensors at the air outlet of the packed
tower assembly.
The concentration sensor would allow for instantaneous reading of the brine concentration in the
bottom tank. It would eliminate the need to take samples and evaporate them later. This
eliminates some of the risk involved with the samples and would eliminate the steps that led to
ruining them. An accurate, instantaneous concentration measurement system would also make it
possible to modify the test while running it to improve the overall results.
15
Home and Office Products. Ecoloblue, 2011. Web. 9 Dec. 2011. <http://www.ecoloblue.com/home-office.html>. 16
Home/Office Products. Air2Water, 2011. Web. 9 Dec. 2011. <http://www.air2water.net/homeoffice.html>. 17
Atmos 28. Atmos H2O, 2011. Web. 9 Dec. 2011. <http://www.atmosh2o.com/docs/59/atmos_28/>.
37
The humidity sensor at the air outlet of the packed tower would allow Team 5 to measure the
change in humidity over the packed tower assembly. This would allow us to determine how
much water the air is losing to the brine while it is traveling down the packed tower. We could
make changes to the rate of air travel or brine travel according to this reading to maximize the
absorption that occurs.
8.5.2. Steelcase Inc.
Improvements that relate directly to the testing performed at Steelcase Inc. include timing the
different cycles more accurately, taking temperature measurements on the condenser coils and
having a concentration sensor in the upper tank.
The timing on the cycles is important to determine when and how often brine should be
circulated up to the top tank and released down into the bottom tank. The timing of when to turn
on the air pump, starting the condenser loop, and when to let the top tank heat is not accurately
known at this time; running more extensive tests to determine the optimal timing of this process
will allow a higher production of water.
Temperature measurements on the condenser coils will give a temperature gradient as the hot
and humid air passes through the condenser loop and will enable the optimal number of coils and
the optimal surface area for heat removal to be determined.
A concentration sensor in the upper tank will allow for an accurate reading of how much water is
in the upper tank. The optimal time to release the brine into the lower tank and when to pump
fresh brine up into the upper tank can be found with a concentration sensor.
Project Design Improvements 9.
Upon finishing the project prototype and performing initial testing, the results obtained are less
than the goal Team 5 set out to accomplish. However, the team recognized several areas of
weakness that could be improved to yield better final results. These improvements can be
separated by whether they would be accomplished with a higher budget or if more time would
make them a reality.
9.1. Financial-Based Improvements
Financially-based improvements would be accomplished by investing more money into the
prototype. Spending more on parts would improve the capabilities of the absorption loop, heating
loop, and condensation loop to maximize water production. Purchasing larger pumps, stronger
fans, more effective heating elements, a more densely packed tower, and improving how these
components are connected would raise the efficiency of the system and enable the atmospheric
water generator to produce more water for a given energy input. Since the working fluid (brine
solution) is a corrosive liquid, making the system with more corrosion resistant materials would
improve product life. Many of the materials in the system are also not very water resistant; the
38
MDF rots when exposed to too much water and the steel frame beams showed oxidation after
only short periods of testing in the humid conditions.
Since the top of the upper tank showed significant warping after being ran at temperature and the
seals for tubes running through the lid leaked frequently, Team 5 would strongly recommend a
custom top tank lid that has the necessary ports formed into it so that tubing connection leaks are
completely eliminated or running the tubes through the side of the tank with high quality seals.
This custom lid would also be heat resistant and provide a way for the safer, more efficient
heating elements to be mounted at the top of the tank. Since heat rises, it might also be a good
idea to implement heaters on the bottom surface of the tank instead of the top. The packed tower
casing and lower tank are also to components that would greatly benefit from custom one-piece
construction. The polycarbonate used for this prototype is expensive and would not have been
used if it was not donated by the Calvin College metal shop. Forming a one-piece lower tank and
a one-piece tower casing and diffuser would also prevent any leaks and alleviate any need to seal
the bottom tank.
The electrical system could also be improved by adding LCD readout screens for the
temperature, humidity, water level, and power consumption monitoring probes. These easier to
understand output screens would also be mounted on a housing for the electrical components.
The housing would not only provide an easier way to mount various sensors and switches to the
frame, but it would also protect all of these components from any brine splashed out of the upper
heating tank. Various switches for the blowers, pumps, and heating elements would also be
installed on this housing to enable easier modulation of those various mechanical components.
9.2. Time-Based Improvements
Time-based improvements are best described as work Team 5 would do if there was enough time
to design and implement them in a second round prototype. Re-wiring the system to keep the
solenoid valves closed when the power is off and re-routing the power to send the most power to
the upper tank in a way that produces the best results of water output.
The heating and condenser loops are both much less efficient than desired; future design work
would involve spending significant time narrowing down exactly how to maximize these
components’ cycle time, physical size, and increase the phase change rate to the fastest point
possible. Adding a third tank that would serve as a buffer tank between the dry, hot, top tank
liquid and the wet, cool, bottom tank liquid would allow the system to cycle faster because hot
brine in the cool brine loop would no longer negatively impact humidity absorption that occurs
there.
To further improve the overall water output and energy consumption numbers, substantial testing
would need to be performed at both of the humidity chambers to ascertain precisely where the
optimum operation points of the system were. Upon determining these optimum points, the
various cycles can be turned on as needed and subsequently operated only when it is beneficial to
39
do so instead of all the time. This optimization process would also open the door for automating
the system and eliminating the need for an operator to flip switches to change cycle operations. It
is unlikely that any of these improvements would increase the water efficiency number (mL/kW-
hr) by a factor of 14 to reach the levels of our competitors, but Team 5 does believe that a little
work will make it easy to meet the goal of 1 liter per day.
Conclusion 10.
After testing in ideal conditions it was found that this wet desiccation system could produce 0.72
liters of water in one day. The system operates at an average of 416 Watts and produces 72.1 mL
per kW-hr. As reference, Ecoloblue 26 produces 1031 mL per kW-hr. This result shows that the
system created by Team 5 is 14 times less efficient at producing water than the competition.
Therefore, after researching, designing, building, and testing a wet desiccation based
atmospheric water generator, Team 5 has determined it is not a practical method of water
generation. In comparison to other atmospheric water generators it is not an improvement on
energy efficiency. While many of aspects of the prototype could be improved, Team 5 doubts
such a system could reach the efficiencies and production rates of current AWGs.
This project has not, however, been a failure; through this project Team 5 hopes to have
promoted development and understanding of wet desiccation technology and has learned
valuable lessons about design and prototyping. Through the extent of this year long design
project each member of Team 5 has had the opportunity to take charge of a portion of this
design. They have cultivated essential skills for managing large projects and meeting important
deadlines. All of this had been done in the context of a technical project which requires
collaboration with others, grounded in design norms and Christian principles. All of these things
will help the members of Team 5 as they move on from Calvin College on the paths that God has
lain before them.
Acknowledgements 11.
Professor Ned Nielsen
Professor Nielsen was the faculty advisor for Team 5. He consistently took opportunities to teach
each member through the different steps of the design process. He encouraged Team 5 to keep a
realistic scope and reminded them that failure happens in the design process, but it isn’t the end
of the road. A change of direction occasionally leads to a better solution.
Professor Michael Harris
Professor Harris proposed the idea of designing an AWG to the senior design class. When Team
5 decided to take on this project, he was valuable in offering preliminary insights and support,
including putting the team in contact with Mr. Steve Beukema. He helped to define an initial
scope and customer. With his help Team 5 was also able to get in contact with several other
people that were willing to help.
40
Mr. Ren Tubergen
Mr. Tubergen is Team 5’s industrial consultant. He provided valuable insight and direction
during the meeting he had with Team 5 as well as through emails. He encouraged the team to
focus on specifying the absorption rate which is the “lynch pin” to designing the system. Without
his honest evaluation we would not have produced the document being set forth.
Professor Douglas Vander Griend
Professor Vander Griend met with Team 5 on several different occasions and he helped design
several experiments through which the absorption rate could be specified. He was instrumental in
procuring some CaCl2 from the Calvin College chemistry department in order for the team to
run the experiments. He set up room in one of the chemistry labs and allowed Team 5 to use
some lab equipment. Without his willingness to help Team 5 would not have been able to
perform these experiments.
Mr. Steve Beukema
Mr. Beukema provided valuable insight into the advantages and disadvantages of filter
technology, particularly as compared to atmospheric water generators. He also offered
encouragement to the team and provided connections to Envirotronics for testing the prototype.
Envirotronics
The company allowed Team 5 to test their prototype in one of their test chambers. This allowed
the team to set a baseline for operation and validate safe operation was possible. Special thanks
go out to Dwayne Botruff and Tim Koenigsknecht who were the contacts at the company.
Steelcase
Steelcase allowed Team 5 to use their humidity chamber which allowed us to test the entire
system and produce water for the first time. A special thanks to Mark Heidmann who connected
us with Steelcase and Kurt Heidmann, Jeff Musculus, and Marty Bender who helped us gain
access to the humidity chamber.
Mr. Phil Jasperes
Phil was considered to be the unofficial 5th
member of Team 5. His expertise was utilized almost
daily and he always offered it willingly. He provided valuable insight on constructing the
prototype as well as insightful ideas as to how the prototype could be improved. We cannot
express our gratitude to Phil enough and we are indebted to him for his assistance.
41
Calvin College Engineering Staff
Bob DeKraker helped Team 5 order all of the parts that were necessary for the prototype. He was
also instrumental in getting the team’s work area set up and established with all of the computers
and technology that was necessary for the project to succeed.
Chuck Holwerda helped Team 5 by offering advice and wisdom whenever he could. He also
helped assist and support many of the tests that were performed.
Michelle Krul was instrumental in organizing and managing all of the teams and the activities
that were involved in this course. She is a master scheduler and organizer.
Rich Huisman, a member of Calvin College’s Chemistry Department, provided Team 5 with
several essential pieces of equipment needed to perform experiments and tests.
Class Advisors
Professor Steven VanderLeest, Professor David Wunder and Professor Wayne Wentzheimer all
helped Team 5 through various assignments and collectively were a vital asset throughout the
senior design class.
42
Appendix 12.
12.1. Competitor Summary
Table 11: Competitor Summary
Model Performance Testing
Conditions
Additional
Features
Cost Water
(L/day)
Power
(W)
Energy
(kJ/L)
Energy
(kJ/kg)
Temp
(C)
Rel.
Hum.
Water
Filters
Hot/
Cold
Water
Dolphin118
$1,799 27.5 500 1570.9 1577.8 30 0.8 5 Y
Ecoloblue
2619
$1,199 26 1050 3489.2 3501.4 28 0.8 11 Y
AirJuicer
401020
$1,499 19 500 2273.7 2281.5 26.7 0.6 5 Y
12.2. Decision Matrix
Table 12: Desiccant Decision Matrix
18
Home and Office Products. Ecoloblue, 2011. Web. 9 Dec. 2011. <http://www.ecoloblue.com/home-office.html>. 19
Home/Office Products. Air2Water, 2011. Web. 9 Dec. 2011. <http://www.air2water.net/homeoffice.html>. 20
Atmos 28. Atmos H2O, 2011. Web. 9 Dec. 2011. <http://www.atmosh2o.com/docs/59/atmos_28/>.
43
12.3. Experiment Setup
Figure 16: Experimental Setup
44
12.4. Experiment Data
Dat
e: 1
2/1/
2011
Exp
eri
me
nt
1
Dat
e o
f
Me
asu
rem
en
ts
Tim
e o
f
Me
asu
rem
en
ts
Tim
e E
lap
sed
(hr)
Ro
om
Tem
p (
°C)
Ro
om
We
t B
ulb
Tem
p (
°C)
Me
asu
red
Mas
s (g
)
Mas
s o
f
Solu
tio
n (
g)
Ch
ange
in
Mas
s (g
)C
on
c. (
w%
)M
eas
ure
d
Mas
s (g
)
Mas
s o
f
Solu
tio
n (
g)
Ch
ange
in
Mas
s (g
)C
on
c. (
w%
)D
ata
coll
ect
ed
by:
12/1
/201
111
:05
0:00
:00
23.5
1846
9.67
60.5
30
0.39
8858
6.90
60.0
30.
000.
4022
EVK
12/1
/201
112
:00
0:55
:00
23.5
1946
9.69
60.5
50.
020.
3987
587.
0460
.17
0.14
0.40
12EV
K
12/1
/201
113
:03
1:58
:00
2312
.546
9.75
60.6
10.
080.
3983
587.
1460
.27
0.24
0.40
06B
TN
12/1
/201
115
:04
3:59
:00
2318
.546
9.83
60.6
90.
160.
3978
587.
3060
.43
0.40
0.39
95R
VB
12/1
/201
115
:58
4:53
:00
2318
469.
8660
.72
0.19
0.39
7658
7.37
60.5
00.
470.
3990
RV
B
12/2
/201
113
:45
26:4
0:00
2315
.547
0.81
61.6
71.
140.
3915
589.
2062
.33
2.30
0.38
73EV
K
12/2
/201
115
:29
28:2
4:00
22.5
13.5
470.
8861
.74
1.21
0.39
1058
9.32
62.4
52.
420.
3866
BTN
12/2
/201
117
:00
29:5
5:00
2218
470.
9361
.79
1.26
0.39
0758
9.43
62.5
62.
530.
3859
RV
B
12/5
/201
112
:00
96:5
5:00
2318
473.
6964
.55
4.02
0.37
4059
4.60
67.7
37.
700.
3564
RV
B
12/5
/201
113
:10
98:0
5:00
2316
473.
7164
.57
4.04
0.37
3959
4.66
67.7
97.
760.
3561
EVK
9.81
5
236.
546
526.
876.
592
409.
14
3.53
5
2.88
7
0.00
633
0.00
422
1.92
1.00
303.
2
235.
8
Surf
ace
Are
a (W
M)
(in
^2)
Surf
ace
Are
a (S
M 1
) (i
n^2
)
Fin
din
g th
e B
rin
e S
olu
tio
n A
bso
rpti
on
Rat
e o
f W
ate
r
Po
t 1
Po
t 2
Dat
a C
oll
ect
ion
Info
rmat
ion
Ro
om
Co
nd
itio
ns
Flu
x R
ate
(SM
) (g
/day
-m2)
Surf
ace
Are
a (W
M)
(m2)
Init
ial C
on
dit
ion
s:
ID S
mal
l Mo
uth
Mu
g (S
M)
(in
)
ID W
ide
Mo
uth
Mu
g (W
M)
(in
)
Mas
s o
f Em
pty
Sm
all M
ou
th M
ug
(g)
Mas
s o
f Em
pty
Wid
e M
ou
th M
ug
(g)
Ro
om
Te
mp
(°C
)
Surf
ace
Are
a (S
M 2
) (i
n^2
)
Flu
x R
ate
(W
M)
(g/d
ay-m
2)
Bri
ne
Ab
sorp
tio
n R
ate
(W
M)
(g/d
ay)
Bri
ne
Ab
sorp
tio
n R
ate
(SM
) (g
/day
)
Surf
ace
Are
a (S
M)
(m2)
45
12.5. EES Calculations for Pressure Dehumification
omega[1]=HumRat(AirH2O,T=28,r=0.8,P=P[1]) dp[1]=DewPoint(AirH2O,T=28,w=omega[1],P=P[1]) omega[2]=HumRat(AirH2O,T=28,D=28,P=P[2]) rh[2]=RelHum(AirH2O,T=28,w=omega[2],P=P[1]) {dp[2]=DewPoint(AirH2O,T=28,w=omega[2],P=P[2])} dp[2]=28[C] P[1]=101 {P[2]=150*convert(psi,kPa)} P[2]=190 rho_water=1000 vol=25 rho_air=density(air,T=20,P=P[1]) eta_compressor=0.80 eta_turb=eta_compressor latent_heat_cond=2257[kJ/kg] e_compress=(P[2]-P[1])*Vol*convert(J,kJ) m_air=rho_air*Vol/convert(m^3,L) m_water=m_air*(omega[1]-omega[2]) q=e_elec/m_water eta_compressor=e_compress/e_elec e_recovered=eta_turb*e_compress e_elec=e_recovered+e_grid q_grid=e_grid/m_water q_recovered=e_recovered/m_water Heat=e_compress+latent_heat_cond*m_water q_competitor=3500[kJ/kg]
Figure 17: EES Display for Pressure Dehumidification
46
12.6. EES Calculations for Desiccation Model
"Model for Peformance of Wet Desiccation AWG" "Specified:" L = 0.3 [m] "Dimension of cubic packing tower" C = 32 "Concentration of the brine solution" m_dot_air = 0.05138 [kg/s] "Mass flow rate of air" T_brine = 30 [C] "Brine temperature in packing tower" "State Points" T[0] = 30 [C] P[0] = 101.325 [kPa] "Assumptions:" m_dot_nominal = 0.00096 [kg/s] "Nominal mass flow rate" j_brine_nominal = 1.9 [kg/m^2*s] "Nominal flux rate" "Calculations:" Factor_SA = (L^3) / ((0.3[m])^3) "Surface area factor" Factor_air = 0.38 + ((0.5[m^2*s/kg]) * (m_dot_air / L^2)) "Air factor" Factor_C = (-1.02 + (0.042 * C)) "Concentration factor" Factor_Brine = (2.6 - (0.053 [1/C] * T_brine)) "Brine factor" m_dot_water = Factor_SA * Factor_air * Factor_C * Factor_Brine * m_dot_nominal m_dot_brine = j_brine_nominal * (L^2) m_dot_waterperday = m_dot_water / convert(s,day) "Mass flow rate of water (kg/day)" "Pump Work Calculations:" h_return = 3 * L "h_return = height from pump to return valve" W_dot_pump_in = m_dot_brine * h_return * g# "Volumetric Flow Rate (air)" rho_air = density(air, T = T[0], P = P[0]) V_dot_air = m_dot_air / rho_air * convert(m^3/s,cfm)
Figure 18: EES Display for Desiccation Model
47
12.7. Complete Project Expenses Table
Table 13: Recorded Project Expenses.
Co
mp
on
en
tM
anu
fact
ure
rP
art
Nu
mb
er
Dis
trib
ute
rSi
zeC
ost
($)
Qu
anti
tySh
ipp
ing
Tota
l Co
st (
$)
Mag
Dri
ve P
um
pM
arch
6305
.201
Rya
nH
erc
o5
GP
M, 1
3 ft
max
he
ad$1
42.8
01
$16.
38$1
59.1
8
Blo
we
r Fa
nLa
sko
(St
anle
y®)
6557
02b
uy.
com
2180
CFM
max
, 8.7
5 lb
s$4
8.96
1$4
.99
$53.
95
Pac
ked
To
we
rC
oo
lin
g To
we
r D
ep
ot
CTD
-ILI
DO
P-1
0C
oo
lin
g To
we
r D
ep
ot
1'x1
'x6'
$40.
211
$26.
89$6
7.10
Air
Blo
we
r P
um
pG
ast
Man
ufa
ctu
rin
gM
OA
-V11
2-A
EM
eta
l Sh
op
1/16
HP
, 115
/110
V, 1
.8/2
.2 A
mp
s$0
.00
1$0
.00
$0.0
0
Co
ils
& C
ou
pli
ng
--23
051
Low
e's
1/4"
x 3
0' c
oil
ed
co
pp
er
tub
ing
$18.
141
$0.0
0$1
8.14
Spra
y N
ozz
les
Orb
it54
020
Low
e's
8' s
pra
y ra
diu
s$1
.80
2$0
.00
$3.6
0
CaC
l 2--
--P
hys
ical
Pla
nt
50 lb
$0.0
01
$0.0
0$0
.00
Low
er
Tan
kM
akro
lon
--M
eta
l Sh
op
1/4"
x 4
' x 6
' $0
.00
1$0
.00
$0.0
0
Bri
ne
Tu
b--
7944
1004
02M
eij
er
14"x
20"x
6"$6
.35
1$0
.00
$6.3
5
Up
pe
r Ta
nk
Insu
lati
on
----
Tran
e2'
x6'x
1"$0
.00
1$0
.00
$0.0
0
Up
pe
r Ta
nk
--47
4971
5002
Me
ije
r5
gal f
ish
tan
k$1
1.12
1$0
.00
$11.
12
Gra
vel
Este
s34
6524
0506
Me
ije
r5
lbs
$4.0
21
$0.0
0$4
.02
Up
pe
r Ta
nk
Stra
ps
--31
7655
Low
e's
12' x
1"
De
sert
Cam
o S
trap
s$1
7.68
1$0
.00
$17.
68
Fitt
ings
U.S
. Pla
stic
Co
rp.
vari
ou
sU
.S. P
last
ic C
orp
.va
rio
us
$12.
161
$9.1
8$2
1.34
Fitt
ings
U.S
. Pla
stic
Co
rp.
vari
ou
sU
.S. P
last
ic C
orp
.va
rio
us
$7.8
01
$6.4
3$1
4.23
Fun
ne
ls--
9559
1Lo
we
's1
pin
t$2
.24
2$0
.00
$4.4
8
Me
sh--
2996
81Lo
we
's36
"x84
"$7
.90
1$0
.00
$7.9
0
Syst
em
Tu
bin
g W
atts
2227
4 &
249
48Lo
we
's3/
4"x3
' an
d 1
/2"x
15' c
lear
vin
yl$1
1.00
1$0
.00
$11.
00
Syst
em
Tu
bin
g W
atts
2227
4Lo
we
's1/
2"x1
0' c
lear
vin
yl$3
.92
1$0
.00
$3.9
2
Syst
em
Tu
bin
g W
atts
24
948
Low
e's
3/4"
ID x
4' c
lear
vin
yl$6
.84
1$0
.00
$6.8
4
Syst
em
Tu
bin
g W
atts
2227
1Lo
we
's3/
8"x1
/4"x
20' c
lear
vin
yl$4
.13
1$0
.00
$4.1
3
Tow
er
Cas
ing
Mak
rolo
n--
Me
tal S
ho
p1/
4" x
4' x
6'
$0.0
01
$0.0
0$0
.00
Syst
em
Fra
me
----
Me
tal S
ho
p1"
x8' s
qu
are
ho
llo
w s
tee
l tu
bin
g$0
.00
8$0
.00
$0.0
0
She
lve
s--
--W
oo
d S
ho
p4'
x6'x
1" m
df
she
et
$0.0
01
$0.0
0$0
.00
Dif
fuse
r--
--M
eta
l Sh
op
4'x4
' ste
el s
he
et
me
tal
$0.0
01
$0.0
0$0
.00
Air
co
lum
n d
uct
ing
----
Me
tal S
ho
p8"
x6' h
oll
ow
pvc
pip
e$0
.00
1$0
.00
$0.0
0
Air
fil
ter
Ene
rgy
Air
e31
2777
Low
e's
1" x
14"
x 2
4" p
leat
ed
air
fil
ter
$0.9
71
$0.0
0$0
.97
Co
lle
cto
r Tu
be
Par
ts--
--Lo
we
's2"
cap
an
d 1
/8"x
11/2
nip
ple
$5.1
91
$0.0
0$5
.19
Ligh
t b
ulb
sG
E--
Low
e's
2x 1
00W
an
d 2
x 40
W$9
.47
1$0
.00
$9.4
7
Ele
ctro
nic
sD
igik
ey
$2
6.34
1$6
.67
$33.
01
Ele
ctro
nic
s--
--R
adio
Sh
ack
--$1
0.78
1$0
.00
$10.
78
GFI
Ou
tle
tC
oo
pe
r24
5864
Low
e's
20 a
mp
, 125
vo
lt$1
6.94
1$0
.00
$16.
94
Ele
ctro
nic
sTi
gerD
ire
ct$1
0.99
1$1
0.00
$20.
99
Tota
l Sys
tem
$512
.33
(Var
iou
s p
arts
are
incl
ud
ed
in t
his
est
imat
e)
(Var
iou
s p
arts
are
incl
ud
ed
in t
his
est
imat
e)
Bu
dge
t fo
r W
ate
r fr
om
Air
(Te
am 5
)
48
12.8. Capacitive Sensor Circuit Reference
Figure 19: PSpice Parametric Capacitive Sweep of Sensor Circuit
CSense
300p 310p 320p 330p 340p 350p 360p 370p 380p 390p 400p
1/period(V(X1:OUTPUT))
11.5K
12.0K
12.5K
13.0K
13.5K
14.0K
14.5K
15.0K
15.5K
16.0K
49
Figure 20: Humidity Sensor Capacitance Response21
Figure 21: Sensor Capacitance Calculation
21
HCH-1000 Series Datasheet, Honeywell International Inc., 2011. Figure 2 in original document.
50
Figure 22: Frequency to Relative Humidity Reference Chart
12.9. Temperature Sensor Circuit Reference
Figure 23: PSpice Voltage Sweep Simulation of Temperature Circuit
0
10
20
30
40
50
60
70
80
90
100
13.5 13.7 13.9 14.1 14.3 14.5 14.7 14.9 15.1 15.3 15.5 15.7 15.9 16.1 16.3 16.5
Re
lati
ve H
um
idit
y (%
)
Measured Frequency (kHz)
V_V4
0.4V 0.6V 0.8V 1.0V 1.2V 1.4V 1.6V 1.8V 2.0V 2.2V 2.4V
V(R2:2)
-12V
-10V
-8V
-6V
-4V
-2V
0V
51
Figure 24: Output Voltage vs. Ambient Temperature22
Table 14: Temperature Probe Calibration Data
Short Lead Medium Lead Long Lead
Temp (C)
Vout (V)
Temp (C)
Vout (V)
Temp (C)
Vout (V)
22.7 0.818 22.7 0.809 22.7 0.834
4.8 0.675 5.8 0.483 5.4 0.545
2.6 0.675 2.6 0.429 2.6 0.456
22
MCP9700/9700A and MCP9701/9701A Datasheet, Microchip Technologies Inc., 2009. Figure 2-16 in original
document.
52
Figure 25: Temperature Probe Calibration Results
y = 0.019x + 0.3764 R² = 0.9997
y = 0.0181x + 0.4268 R² = 0.9904
y = 0.0075x + 0.6479 R² = 0.99
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
0 5 10 15 20 25
Raw
Ou
tpu
t V
olt
age
(V
)
Temperature (C)
Medium Lead
Long Lead
Short Lead
Linear (Medium Lead)
Linear (Long Lead)
Linear (Short Lead)
53
Figure 26: Temperature Circuit Calibration Mathcad Sheet (Screenshot)
54
12.10. Envirotronics Validation Certificate
55