An experimental study on an

evaporative cooler for hot rural areas

Katarina Gustafsson
Hanna Simson

Bachelor of Science Thesis

KTH School of Industrial Engineering and Management
Energy Technology EGI-2016
SE-100 44 STOCKHOLM
  
    Bachelor  of  Science  Thesis  EGI-­‐2016  
 
An  experimental  study  of  an  evaporative  cooler  for  hot  
rural  areas
 

     
    Katarina  Gustafsson  
Hanna  Simson  
Approved   Examiner   Supervisor  
  Catharina  Erlich   Eyerusalem  Birru  
  Commissioner   Contact  person  
Engineers  without  borders   Sebastian  Haglund  El  Gaidi  

Acknowledgments  
We would like to thank our supervisor Eyerusalem Birru and our contact person at Engineers
without borders, Sebastian El Gaidi, who made this project possible and for all the great help
and feedback they have provided. We would also like to thank Catharina Erlich, Peter
Kjaerbo and the Department of Energy, CETER, for feedback and advice. A special thanks to
Peter Hill who helped us with the lab setup and with some great advice. Last but not least we
would like to thank Livio Jaconelli and Kim Jaconelli Palm for good cooperation between the
groups.

1
 Abstract  
In developing countries about 40 % of the food-waste is due to post-harvest losses, such as
improper storage. The Zeer-pot is an evaporative cooler, which cools the inside by convective
heat transfer, and can be used to keep fruits and vegetables fresh longer. This is typically
convenient in hot rural areas without access to electricity and is more efficient in non-humid
areas.

This study will investigate if there is a correlation between the temperature decrease inside the
pot and the wind velocity, how the efficiency of the Zeer-pot is affected by hanging it in the
air to additionally expose the underside to the airflow and also how the efficiency is affected
from glazing of the inner pot, to prevent the food from getting damaged from high humidity.
It will also consider the feasibility of combining the Zeer-pot with a solar dryer, also to
improve its efficiency.

The evaporation is increased by higher wind velocity due to forced convection. A solar dryer
can create an airflow when the heated air rises along the surface of the solar collector and
creates a temperature difference between the upper and the lower part. A design for a solar
dryer that could be appropriately integrated with a Zeer-pot to achieve a greater airflow
around it is modelled in CAD and presented.

The tests on the pots took place in a climate chamber where the ambient temperature was
controllable. In the climate chamber a fan and a dehumidifier was installed in order to create
wanted conditions. One pot was tested only for wind velocities and the other only for the
cases of the pot hanging in the air and being glazed on the inside. A reference case was
designed and tested for the second pot in order to compare the glazed and hanging pot in the
same conditions.

The relative humidity was not controllable in this setup, and thereof a way to compare these
results was to calculate the final temperature the pot achieved relative to the lowest possible
theoretical temperature, the wet bulb temperature.

For the first pot an almost linear correlation between the time it took to reach the final
temperature depending on the wind velocities could be observed, apart from two values. A
rather nice coherent curve, also apart from two values, was found for how close to the wet
bulb temperature the final temperature was depending on the wind velocity. For the second
pot the cooling capacity was enhanced for both the hanging construction and the glazed pot.
For the hanging pot this was expected, but for the glazed one it was not.

If a solar dryer is combined with the Zeer-pot, a wind velocity around 3-3.5 m/s is guaranteed
to improve the Zeer-pots cooling capacity. A lower wind velocity could probably make a
large difference too, but the experiments in this study is insufficient to make any conclusions.

2
 Sammanfattning  
I utvecklingsländer beror cirka 40 % av matavfallet på förluster efter skörd, till exempel
felaktig förvaring. En Zeer-pot är en evaporativ kylare som kyls med hjälp av konvektion och
kan användas för att öka livslängden på frukter och grönsaker. Den är användbar i områden
utan tillgång till elektricitet där klimatet är varmt och behovet för kylning är stort. Kyleffekten
förbättras i ett klimat med låg luftfuktighet.

Denna studie kommer att undersöka om det finns ett samband, mellan temperatursänkningen
och vindhastigheten. Ett sätt att förbättra kyleffekten skulle kunna vara att konstruera någon
form av stöd för att göra det möjligt för krukan att hänga i luften, detta för att även undersidan
ska utsättas för luftflödet. För att inte låta fukten som skapas vid evaporationen komma in i
innerkrukan där maten förvaras kan innersidan glaseras. Hur mycket dessa justeringar
kommer att påverka kylkapaciteten undersöks också i denna studie.

Krukan kyls med hjälp av evaporation av vatten och denna ökar vid en högre vindhastighet på
grund av påtvingad konvektion. Ett sätt att skapa ett större luftflöde runt krukan är att
integrera en soltork med den så kallade Zeer-poten. Soltorken skapar ett luftflöde då en
temperaturskillnad mellan den övre och den undre delen får värmen att stiga upp från ytan av
solfångaren. En hypotetisk modell av en lämplig soltork modelleras i CAD och presenteras.

För att se en korrelation för hur ökade vindhastigheter påverkar temperatursänkningen i

grader och hur fort sluttemperaturen uppnås gjordes tester på samma kruka i så i övrigt
konstanta förhållanden som praktiskt var möjligt.

Testen för de olika krukorna ägde rum i en klimatkammare där luftens temperatur gick att
kontrollera. En fläkt och en avfuktare var installerade. Den ena krukan testades för fallet med
olika vindhastigheter och den andra för den upphängda krukan och för glaseringen på insidan.
Ett referensfall gjordes även för att jämföra den andra krukans två fall separat.

Den relativa fuktigheten gick inte att kontrollera i klimat-kammaren och därav fick resultaten
jämföras genom en framräknad procentsats. Procentsatsen visade hur långt testets
sluttemperatur hade sjunkit i förhållande till den våta temperaturen.

En nästan linjär korrelation, med undantag för två värden, mellan tiden det tog att nå
sluttemperaturen i förhållande till vindhastigheten kunde observeras. Ett tydligt samband,
också undantaget från två värden, kunde observeras mellan den våta temperaturen och den
uppnådda temperaturen för olika hastigheter. För den andra krukan ökade kylkapaciteten för
både den hängande konstruktionen samt den glaserade krukan. Det var väntat för den
hängande konstruktionen, men inte för den glaserade.

Om en soltork kombineras med Zeer-poten i syfte att märkbart förbättra dess kylkapacitet,
skulle ett luftflöde med en vindhastighet kring 3-3,5 m/s garanterat fungera. Lägre
vindhastigheter lär även de kunna påverka krukan nästan lika mycket, men tyvärr är resultaten
från experimenten inte tillräckliga för att dra några slutsatser om detta.

3
Table  of  Content  
ACKNOWLEDGMENTS .......................................................................................................................... 1  
ABSTRACT............................................................................................................................................ 2  
SAMMANFATTNING .............................................................................................................................. 3  
TABLE OF FIGURES ............................................................................................................................... 5  
TABLE OF TABLES ................................................................................................................................ 5  
NOMENCLATURE.................................................................................................................................. 6  
1.   INTRODUCTION ......................................................................................................................... 7  
1.1.   PREVIOUS STUDIES ................................................................................................................. 8  
2.   OBJECTIVES ............................................................................................................................... 9  
2.1 PROBLEM FORMULATION............................................................................................................ 9  
2.2 AIM OF STUDY .............................................................................................................................. 9  
3.   LITERATURE SURVEY ........................................................................................................... 10  
3.1.   OTHER AREAS OF APPLICATIONS FOR EVAPORATIVE COOLING............................................ 10  
3.2.   HEAT TRANSFER ................................................................................................................... 10  
3.2.1.   Conduction ................................................................................................................... 10  
3.2.2.   Convection ................................................................................................................... 11  
3.2.3.   Radiation...................................................................................................................... 12  
3.2.4.   Total amount of energy emitted ................................................................................... 13  
3.3.   RELATIVE HUMIDITY ............................................................................................................ 13  
3.4.   THE MOLLIER-DIAGRAM ...................................................................................................... 13  
3.5.   WET SAND – AN EVAPORATOR AND INSULATOR .................................................................. 13  
3.6.   ALTERNATIVES TO SAND ...................................................................................................... 14  
3.7.   SIMULATIONS OF THE ZEER-POT .......................................................................................... 14  
3.8.   SOLAR DRYERS ..................................................................................................................... 14  
4.   DIFFERENT CONCEPTS OF SOLAR DRYERS INTEGRATED WITH THE ZEER-POT
16  
4.1.   PLACEMENT IN FRONT OF THE SOLAR COLLECTOR T............................................................ 16  
4.2.   OTHER CONCEPTS OF SOLAR COLLECTORS ........................................................................... 18  
5.   MARKET ANALYSIS................................................................................................................ 20  
5.1.   BURKINA FASO ..................................................................................................................... 21  
5.1.1.   Climate ......................................................................................................................... 22  
5.1.2.   Business plan ............................................................................................................... 23  
6.   METHODOLOGY ...................................................................................................................... 24  
6.1.   EXPERIMENTS AND THEIR SETUP .......................................................................................... 24  
7.   RESULTS..................................................................................................................................... 26  
7.1.   RESULTS FOR THE WIND VELOCITY TESTS ............................................................................ 26  
7.2.   RESULTS FOR THE HANGING CONSTRUCTION ....................................................................... 29  
7.3.   RESULTS FOR THE GLAZED POT ............................................................................................ 31  
8.   DISCUSSION AND CONCLUSIONS ...................................................................................... 32  
8.1.   GENERAL SOURCES OF ERROR .............................................................................................. 33  
8.2.   SPECIFIC SOURCES OF ERROR ............................................................................................... 35  
8.3.   CONCLUSIONS ....................................................................................................................... 36  
8.4.   RECOMMENDED FUTURE WORK............................................................................................ 37  
9.   REFERENCES ............................................................................................................................ 38  
10.   APPENDIX .............................................................................................................................. 43  

4
 Table  of  figures  
 
FIGURE  1;  A  PRACTICAL  USE  OF  THE  ZEER-­‐‑POT  [7].  ...............................................................................................................................  8  
FIGURE  2;  AN  ILLUSTRATION  OF  A  CONSTRUCTION  EXPOSING  THE  ZEER-­‐‑POTS  UNDERSIDE  TO  THE  AIRFLOW.  ............................  9  
FIGURE  3;  AN  ILLUSTRATION  OF  THE  PRINCIPLE  OF  COOLING  BY  CONVECTION  FOR  THE  ZEER-­‐‑POT  [14].  .................................  11  
FIGURE  4;  AN  ILLUSTRATION  OF  HOW  THE  AIRFLOW  IS  MOVING  THROUGH  THE  SOLAR  DRYER.  THE  BLUE  ARROW  IS  
REPRESENTING  COLD  AIR  AND  THE  RED  ARROW  IS  REPRESENTING  HEATED  AIR.  ................................................................  15  
FIGURE  5;  ILLUSTRATIONS  OF  A  PROPOSED  SOLAR  DRYER  MODEL  DESIGN  ADAPTED  FOR  INTEGRATION  WITH  A  ZEER-­‐‑POT.  
THE  BLUE  ARROWS  ARE  REPRESENTING  COLD  AIR  COMING  IN  AND  THE  RED  ARROWS  IS  REPRESENTING  HEATED  AIR  
GOING  OUT.  ......................................................................................................................................................................................  17  
FIGURE  6;  THE  SAME  SOLAR  DRYER  AS  FIGURE  5  FROM  ANOTHER  VIEW.  .........................................................................................  17  
FIGURE  7;  A  SOLAR  CHIMNEY  AND  ZEER-­‐‑POT  INTEGRATED  WITH  A  HOUSE.  ....................................................................................  18  
FIGURE  8;  AREA  IN  FOCUS  [28].  ..............................................................................................................................................................  20  
FIGURE  9;  BURKINA  FASO  DIVIDED  INTO  DIFFERENT  CLIMATE  ZONES  BY  THE  KÖPPEN  CLIMATE  CLASSIFICATION  [29].  .......  22  
FIGURE  10;  THE  LAB  SETUP  WITH  THE  DEHUMIDIFIER  TO  THE  LEFT  AND  THE  POT  WITH  THE  FAN  BEHIND  IT  TO  THE  RIGHT.25  
FIGURE  11;  THE  DIFFERENT  TIMES  TO  REACH  THE  FINAL  TEMPERATURE  DEPENDING  ON  THE  WIND  VELOCITY.  .....................  27  
FIGURE  12;  THE  TEMPERATURE  DECREASE  IN  PERCENTAGE  OF  THE  LOWEST  THEORETICAL  TEMPERATURE  FOR  DIFFERENT  
WIND  VELOCITIES.  ..........................................................................................................................................................................  28  
FIGURE  13;  A  GRAPH  OF  THE  RESULTS  FOR  THE  REFERENCE  CASE.  ...................................................................................................  29  
FIGURE  14;  A  GRAPH  FOR  THE  CASE  OF  THE  POT  HANGING  IN  THE  AIR.  ............................................................................................  30  
FIGURE  15;  A  GRAPH  FOR  THE  CASE  OF  THE  POT  GLAZED  ON  THE  INSIDE.  .......................................................................................  31  

Table  of  tables  

TABLE  1;  POPULATION  AND  ACCESS  TO  ELECTRICITY  [4].  ..................................................................................................................  21  
TABLE  2;  FINAL  TEMPERATURE  AND  COOLING  TIME  DEPENDING  ON  WIND  VELOCITY.  ..................................................................  26  
TABLE  3;  RELEVANT  DATA  COMPARED  FOR  THE  CASES  OF  DIFFERENT  WIND  VELOCITIES.  ............................................................  27  
TABLE  4;  RELEVANT  DATA  FOR  THE  REFERENCE  CASE.  .......................................................................................................................  29  
TABLE  5;  RELEVANT  DATA  FOR  THE  CASE  OF  THE  ZEER-­‐‑POT  HANGING  IN  THE  AIR.  .......................................................................  30  
TABLE  6;  RELEVANT  DATA  FOR  THE  CASE  OF  THE  GLAZED  ZEER-­‐‑POT.  .............................................................................................  31  
TABLE  7;  HOW  THE  RELATIVE  HUMIDITY  VARIED  IN  EACH  TEST.  .......................................................................................................  43  

5
 Nomenclature  
Symbol Definition Unit
𝑄     Heat transfer [W]
ℎ     Convection heat transfer coefficient [W/m2·K]
𝑘     Thermal conductivity [W/m·K]
𝐴     Area [m2]
∆𝑥     Length [m]
𝑇     Temperature [K]
𝑇( − 𝑇*     Temperature difference [K]
𝐿,     Characteristic length for geometry [m]  
𝑁𝑢     Nusselt number [-]
𝑅𝑒     Reynolds number [-]
𝑃𝑟     Prandtls number [-]
𝜌     Clay density [kg/m3]
𝑉     Velocity of the air [m/s]
𝜀     Emissivity of the surface [-]
𝜎     Stefan Boltzman constant [W/m2·K4]
𝑐8     Specific heat [J/kg·K]
𝜇     Dynamic viscosity [kg/m·s]
𝜑     Relative humidity [-]
𝑝     Pressure [Pa]  
   
 
Subscript Explanation
𝑐𝑜𝑛𝑣     Convection
𝑟𝑎𝑑 Radiation
𝑡𝑜𝑡 Total
𝑒𝑣𝑎𝑝 Evaporation
𝑐𝑜𝑛𝑑 Conduction
𝑠 Surface
∞ Ambient air
𝑠𝑢𝑟𝑟 Surroundings
𝑎𝑣𝑒𝑟𝑎𝑔𝑒 Average
𝑤𝑒𝑡𝑏𝑢𝑙𝑏 Wet bulb temperature
𝑠𝑎𝑡 Saturation pressure
𝑝𝑎𝑟𝑡 Partial pressure

6
 1.   Introduction  
Hunger is a global problem and about 795 million people in the world do not have enough
food to live a normal lifestyle. The highest hunger prevalence is in Sub- Saharan Africa,
where about one fourth of the population is undernourished [1].

About one third of the world’s total food production is wasted before it is consumed, this is
referring to all the spoilage from the initial stages, to production and to the private consumer’s
waste. This means that recovering just half of that wastage could feed all the people in the
world. In industrialized countries most of these food losses occur at a consumer level, while
in developing countries about 40 % of the waste is due to post-harvest losses, such as
improper storage [2]. In areas like Sub-Saharan Africa the burden of food losses falls on the
smallholder farmers and families [3].

A majority of people in western Africa, the area that is investigated in the current study, have
no access to electricity [4]. Thus, refrigerated storage is presently not an option, resulting in
large food spoilage and economical as well as health problems.

The Zeer-pot is an evaporative cooler, which can be used to keep fruits and vegetables fresh
longer. This is typically convenient in hot rural areas without access to electricity and is more
efficient in non-humid areas. There are records of the usage of this method that dates back
several centuries in the Middle East and Africa. The Nigerian Mohammed Bah Abba
reinvented the method in the 1990s [5].

The evaporative cooling method is used to cool the inside of a pot where food can be stored.
The Zeer-pot is constructed out of two clay pots, where the outer pot is of a larger size than
the inner pot and the gap between them is filled with wet sand, see Figure 1. Evaporation of
water drives the heat from the inside to the outer surface of the pot and then out into the air;
this is the same principle as when humans cool by sweating. The wet sand also acts as an
insulator as the inner pot is chilled. The evaporation will be more efficient with a light breeze
and a dry environment [6].

This study aims to investigate the feasibility of different models of Zeer-pots and to
experimentally determine whether higher wind velocities have potential on improving the
cooling capacity and if so, by how much. A similar study by Sebastian Haglund El Gaidi [6],
where the physical properties of the Zeer-pot were investigated, has shown that an increased
airflow around the Zeer-pot will lower the storage temperatures significantly. Hence, this
study will examine if the hypothesis that an increased airflow will lower the cooling capacity
can be validated and also to propose a conceptual design for combining the Zeer-pot with a
solar dryer in order to increase the airflow. An appropriate business plan and marketing
strategy for the most suitable country considering the climate, in this study Burkina Faso, is
required in order to adapt the concept to the areas in need.

7
 Figure 1; a practical use of the Zeer-pot [7].

1.1.   Previous  studies  

The principle of using evaporative cooling as a cooling mechanism with clay materials dates
back several centuries. The method of the pot-in-pot-evaporator was rediscovered by the
Nigerian Mohammed Bah Abba in 1990s [5]. Since then, a few laboratory studies on the so
called Zeer-pot have been made.

Practical women’s association have tested Mohammed Bah Abbas design and investigated
how the Zeer-pot can improve shelf-life of vegetables. For example, a significant increase of
the shelf-life of tomatoes from 2 days to 20 days was observed [8].

In laboratory conditions the Nigerian Victor Aimiuwu managed to cool two claypots, using
water instead of sand to fill the gap between the pots, by 14 degrees and by that confirming
the Zeer-pots good cooling capacity [5].

Experimental studies have not previously been made on how the wind velocity affects the
cooling capacity. Sebastian El Gaidi have by simulations using different wind velocity found
that this and changing some other parameters, like using materials with lower thermal
conductivity, can improve the cooling capacity [6].

8
 2.   Objectives  
In this section the problem this study aims to solve is shortly presented and it also expresses
the goals for the project.

2.1          Problem  formulation  

As mentioned, the difficulties with hunger and storage opportunities due to lack of access to
electricity in developing countries is the problem aspired to find a solution to in this project.
One proposition for a solution is a Zeer-pot combined with a solar dryer to create a
multipurpose design, since increasing the wind velocity around the pots will increase its
efficiency and thereof decrease the final temperature of the inner pot and also drying food at
the same time. Both will make the food last longer.

2.2      Aim  of  study  

The goals for this study are to:
•   Investigate if there is a correlation between temperature differences and the wind
velocity.
•   See how a hanging construction, to additionally expose the underside to the airflow,
will affect the cooling capacity. An example of a potential construction for this is
illustrated in Figure 2.
•   See how glazing of the inner pot to protect the food from moist and high humidity,
will affect the cooling capacity.
•   Investigate if the results shows how great of a wind velocity is needed to make a
significant change in the cooling capacity.
•   Present a hypothetic model of a solar dryer to combine with the Zeer-pot.

Figure 2; a construction exposing the Zeer-pots underside to the airflow.

9
 3.   Literature  survey  
This part presents the theoretical background to the study, such as heat transfer mechanisms,
relative humidity and solar dryer principles.

3.1.   Areas  of  applications  for  evaporative  cooling  

The definition for evaporative cooling is a temperature decrease due to heat being transferred
into the air as water evaporates. [10]. There are other more established usages of the
evaporative cooling techniques than the Zeer-pot for different needs of cooling. Some
traditional air conditioners use this type of process, although they need electricity to operate
the pumps [10].

Other types of evaporative coolers which uses less energy to operate, is for example some
buildings that uses sprayed water mists. This cools the inside air as the water evaporates [11].
Another technique is the roof pond, which is an open water tank on a building’s roof, which
cools the building’s exterior and consequently keeps the inside in a comfortable temperature
as the water evaporates [10].

3.2.   Heat  Transfer  

To calculate how much heat that leaves a body there is three different types of heat transfer
mechanisms that must be considered. Those mechanisms are: conduction, convection and
radiation.

3.2.1.   Conduction  

Conduction is when a more energetic particle, for example one with higher temperature,
transfers energy to a less energetic particle (lower temperature) through interactions between
the two. This is due to collisions and diffusion of the molecules in fluids and gases, while in
solids it is mainly due to vibrations and exchange of free electrons. The amount of heat
transferred through conduction is written:
LM NLO
𝑄,HIJ = 𝑘𝐴  [𝑊], (1)
∆P

where k is the thermal conductivity [W/m·K] which is a measure of how much heat the
material can conduct, A is the area [m2] of the material over which the heat is transferred
through, 𝑇( − 𝑇* is the temperature difference across the material [K] and ∆𝑥 is the length of
the material in which the heat is transferred [m] [12].

10
 3.2.2.   Convection  

Convection is the one that has the greatest influence of the three, in the given situation with a
pot placed in the shade, and it includes both heat conduction and bulk fluid motion.

Basically there is two types of convection; forced and natural convection. Forced convection
is typically illustrated with a fluid or gas that is forced to flow over a surface due to external
means, for example caused by a fan or the wind. Natural, also called free, convection is when
the fluid motion is caused by buoyancy forces that are induced by differences in density and
temperature. Usually the forced convection is more effective in transferring heat than the
natural convection. For example, a common way to test in what direction the wind is blowing
is to lick your finger and raise it into the air, away from your body. The side of the finger on
which the wind is blowing will feel a lot colder than the other side. This is because there is
forced convection on the windy side and only natural convection on the other.

If the air over a surface is in motion, it will increase the heat transfer from the surface. This is
because the motion will replace the already heated air close to the surface with cooler air.
Thereby you get a greater temperature difference between the body and the air and more heat
can be transferred.

The same principle goes for mass transfer of for example water on a surface into the air. The
water strives to be evenly distributed throughout the mediums and therefor diffuses into the
air, as long as the relative humidity is lower there. Because of this, the water will also keep
spreading evenly throughout the material if it’s porous enough. The outer surface of a Zeer-
pot for example will be kept at a relatively moist state and the evaporation will continue until
there is no water left in the material or until the air has an equal or higher relative humidity.
As water evaporates from a surface it requires energy or heat to make the water transform
from a liquid to a gas. This energy is taken from the water still on the surface of and inside the
body that it is leaving, making the water decrease in temperature and thereof cooling the
body.

Both of these features are very important in the role of cooling a body or an object, see Figure
3. A greater temperature difference between the air and the surface, but also a wet surface in
contact with dryer air, will result in more heat being transferred between the two [13].

Figure 3; an illustration of the principle of cooling by convection for the Zeer-pot [14].

11
The amount of energy emitted from a body due to convection is

𝑄,HIU = ℎ𝐴V (𝑇V − 𝑇X ) [W], (2)

where 𝐴V is the area of the surface of which the convection happens [m2], 𝑇V is the
temperature of the surface [K], 𝑇X is the ambient air temperature [K] and ℎ is the convection
heat transfer coefficient [W/m2K]. To find ℎ a formula for Nusselts number is used:
Z[\
𝑁𝑢 = , (3)
]

where 𝐿, is the characteristic length for the geometry [m], defined as the volume of the object
divided by its surface area. The Nusselt number is depending on both whether the flow of air
is laminar or turbulent around the body and on Reynolds and Prandtls numbers. The two later
are calculated as
^_[\
𝑅𝑒 = (4)
`

and
`,a
𝑃𝑟 = , (5)
]

where 𝜌 is the density [kg/m3], 𝑉 is the velocity of the air [m/s], 𝜇 is the dynamic viscosity
[kg/m·s] and 𝑐8 is the specific heat [J/kg·K] [15].

The air can contain different percentages of relative humidity, which will also affect the
amount of heat being transferred from the body. Read more about this in the section 3.3
Relative humidity.

3.2.3.   Radiation  

Radiation transfers heat through emission of electromagnetic waves. These waves bring
energy with them from the warmer body to a colder body or the surrounding. Unlike
conduction and convection, the radiation heat transfer does not depend on having an
intervening medium present. The rate of radiation heat transfer is explained by the Stefan-
Boltzmann law:

𝑄bcJ = 𝜀𝜎𝐴V (𝑇Vd − 𝑇Vebb

d
) [W], (6)

where 𝜀 is the emissivity of the surface (0 ≤ 𝜀 ≤ 1), 𝜎 is the Stefan-Boltzmann constant

which is equal to 5.670·10-8 [W/m2·K4], 𝑇V is the temperature on the surface [K] and 𝑇Vebb is
the temperature of the surrounding [K]. Radiation is not limited to transferring heat from a hot
surface to a colder surrounding, if the surroundings are warmer than the surface, the 𝑄bcJ will
attain a negative value and thereby the surface will absorb energy instead of emitting it [16].

12
 3.2.4.   Total  amount  of  energy  emitted  

From these three types of heat transfer the total amount of energy emitted from the body to
the air is calculated as

𝑄iHi = 𝑄jUc8 +𝑄,HIU [W], (7)

since 𝑄,HIJ is included in 𝑄,HIU [16].

3.3.   Relative  humidity  

Relative humidity is a measure of how saturated the air is with water vapor. The definition for
relative humidity is
8almn
𝜑= , (8)
8oln

where 𝑝8cbi is the water vapor’s partial air pressure and 𝑝Vci is the water vapor’s saturation
pressure. The water vapor is on the verge of becoming a fluid when 𝜑 = 1, which is when the
phenomena of fog occurs. If the temperature of the air increases, with the same amount of
water in the air, the saturation pressure will get larger, while the partial pressure remains the
same and the relative humidity will attain a lower value [17].

The lower the value of the relative humidity, the more water can diffuse into the air before
reaching saturation. This means that if the air has a relative humidity of 𝜑 = 0.1, it can absorb
a whole lot more water in a shorter period of time than if 𝜑 = 0.95 and is almost saturated.

3.4.   The  Mollier-­‐diagram  

To find the lowest possible temperature theoretically, called the wet bulb temperature [18],
for the inner pot a Mollier-diagram is used, see Appendix 3 [19]. The diagram describes the
relations between the relative humidity, the ambient temperature and the wet bulb
temperature, which is the lowest achievable temperature due to evaporation. Therefor if the
ambient temperature and relative humidity is known in the area, the theoretical best
temperature inside the Zeer-pot can be found.

3.5.   Wet  sand  –  an  evaporator  and  insulator  

The decision to use sand between the pots is due to its properties, both as dry and wet. While
the sand is saturated with water it offers a thermal conductivity of around 2-4 𝑊/𝑚 ∙ 𝐾,
meaning it conducts heat better than pure water, conducting about 0.58  𝑊 /𝑚 ∙ 𝐾. As the sand
is then dried out the thermal conductivity is lowered to about 0.15-0.25 𝑊/𝑚 ∙ 𝐾 [20]. This
property is useful in context with the Zeer-pot. The higher thermal conductivity when the
sand is wet makes it easier for the heat to transfer throughout the sand, resulting in a faster
temperature decrease inside the Zeer-pot. As all the water is evaporated, the dry sands lower

13
thermal conductivity will instead make it work as an insulator, keeping the pots cool for a
longer amount of time.

Even though the heat is continuously removed, there is a limit to how cold the inner pot can
get. This minimum temperature is called the wet bulb temperature as mentioned in section
3.4. The Mollier-diagram. The time it takes to lower the inner temperature and also how low
it sinks depends on the factors mentioned above: conduction, convection and radiation. How
rapidly the temperature is lowered depends on how much water there is in the sand as well as
the temperature of the water. It also depends on how thick the pots are and how much food
that is stored inside the Zeer-pot, since this needs to be cooled too.

3.6.   Alternatives  to  sand  

Studies are made using other insulating materials than sand between the pots, and there are
multiple choices that will provide an equivalent or better efficiency for the Zeer-pot. One of
those alternatives is charcoal. Charcoal can lower the temperature one or a few additional
degrees in an empty Zeer-pot compared to sand and about one degree when the Zeer-pot is
loaded with 500 grams of tomatoes [21]. Quite a lot of charcoal is produced in Burkina Faso,
which is the area chosen to focus on in this study, see 5. Market analysis, but the availability
seems to be mainly limited to households in the cities, meaning it is not a lot of rural
households in possession of it [22][23]. The sand will therefor be the most accessable of the
two.

3.7.   Simulations  of  the  Zeer-­‐pot  

By simulations given to us from a cooperative group [24], the temperature of the inner pot is
supposed to reach 28.8 °C in an ambient temperature of 40 °C, a relative humidity of 39 %
and a wind velocity of 3 m/s. This results in a decrease of 90 % relative to the wet bulb
temperature, calculated as described in equation 10 in section 6. Methodology. The
parameters in the simulations are assumed optimal for the pot and thereof result in a high
efficiency value. This will be compared to the experimental results in 8. Discussion and
conclusions.

3.8.   Solar  dryers  

Drying techniques in its simplest form has been used to preserve food for centuries. Farmers
in agricultural communities can preserve food for a longer time and instead of having to sell
the food directly as harvested they could sell on customer’s demand. The storage area can also
be decreased with this technique as the liquid evaporates from the food and the volume
decreases [25].

Sun drying means that the food is spread out in the direct sunlight and open air and dries by
direct exposure to solar radiation. The simplicity of this process is its biggest advantage as it
can be set up anywhere, as long as the temperature is high enough, and with very limited
resources. The disadvantage is that it is dependent on weather conditions and that the food
could easily be contaminated as it is exposed to the surrounding environment with potential
bacteria, dust, insects and humidity [26].

14
As opposed to sun drying, solar drying uses indirect solar radiation. The food lies in a
separate container, where some sort of black surface collects the solar radiation and converts
the light to heat more efficient than other surfaces. A clear solid is used to isolate the heat
from the outside air, but still allowing the solar radiation to pass through the solid. This makes
the solar dryer dependent only on the solar intake and not on other environmental conditions.
The food is dried by natural or forced convection by the trapped heat. The heat that rises up
from the surface creates a flow of air from the lower to the upper part of the solar collector
[27]. This is illustrated in Figure 4, which is a version where there are two separated airflows.
This is explained more in section 4. A model draft of a solar dryer design integrated with the
Zeer-pot.

Figure 4; an illustration of how the airflow is moving through the solar dryer. The blue
arrow is representing cold air and the red arrow is representing heated air.

There are a few different categories of solar dryers; natural convection, forced convection and
tunnel dryers. As this study aims to investigate areas with no electricity, a natural convection
dryer is of interest, as other solar dryer systems require some energy source in order to force
the airflow [25].

15
 4.   Different  concepts  of  solar  dryers  
integrated  with  the  Zeer-­‐pot  
The naturally attained airflow from the upper and lower part of the collector is intended to be
used around the Zeer-pot for a more efficient cooling. The pot can be placed both before and
after the solar collector, either way the pots will be exposed to a constant airflow. A natural
convection dryer can be modelled in different ways, for example the solar collector heats the
air, which will then be led to the food cabinet enabling the food drying. For the purpose of
increasing the Zeer-pots cooling capacity, it is desirable to avoid that the airflow obtained
from the solar dryer attains a higher relative humidity, which will happen once in contact with
the drying food. This can be done by either placing the Zeer-pot in the inlet of the air or by
another proposed design described in the next section.

4.1.   Placement  in  front  of  the  solar  collector    

To minimize the problem with the humid air, an idea is to place the food cabinet under the
solar collector. By that design, the air can go two ways. One is through the food cabinet that
will be heated with the solar collector and one is through the solar collector, this is seen in
Figure 5 and Figure 6. The air in the food cabinet will attain a higher relative humidity from
drying the food, while the air going through the solar collector will not. The two airstreams
will then mix up in a pipe, marked with the red arrow closest to the pot in Figure 5 and
Figure 6. The mix will prevent the air from getting too humid and the pipe will create a
breeze directed at the Zeer-pot.

16
 Figure 5; illustrations of a proposed solar dryer model design adapted for integration
with a Zeer-pot. The blue arrows are representing cold air coming in and the red arrows
is representing heated air going out.

Figure 6; the same solar dryer as figure 5 from another view.

The advantage with this model is that the air is dryer reaching the Zeer-pot compared to the
air passing through the food cabinet. The obtained light breeze created by the solar collector
will improve the cooling capacity of the pot as well as the constant airflow will make the
process of drying food a lot shorter. Additionally, the solar collector is easy to build and have
a simple geometry.

By placing the food cabinet under the solar collector in order to avoid the problem with the
humid air reaching the Zeer-pot, could decrease the drying efficiency due to less exposure to
solar radiation. Other disadvantages are that the solar collector needs to be placed in the sun
to work, for example on the roof, but that the Zeer-pot works best placed in the shade,
preferably inside the house. This means that the airflow created by the solar collector will

17
have to be led through a pipe to the Zeer-pot, standing at a distance. Making this model work
in practice is not problem-free, for example is losing too much wind velocity due to friction in
the pipe a major problem. To prevent longer transportations of the air, one alternative is to
place the pot outside, closer to the solar collector, and build some kind of roof over it to keep
it in the shade but still exposing the solar collector to the sun. Since the air is heated and then
led into the house, it results in that the house will also get somewhat heated depending on how
much the air has cooled down before reaching the inside.

4.2.   Other  concepts  of  solar  collectors  

Placing the Zeer-pot at the outlet of the solar dryer is one way to combine the two to achieve a
greater efficiency. Another way is to place the pot at the inlet and instead have an airflow
created by suction. A type of solar collector most suited for this is usually called a solar
chimney. The solar chimney can be viewed as a solar collector in the shape of a chimney. At
the bottom of the chimney there is an inlet and at the top there is an outlet. The solar chimney
is heated by the sun, creating the warmed air to rise up through the chimney and by this
creating an airflow. In the case with the Zeer-pot, this solution could be more easily
applicable since it could easily be built in connection to the house. It could also help cool
down the house since the flow created can make cool air from the side of the house in the
shadow come in and replace the heated air inside. A schematic illustration of this concept is
shown in Figure 7.

Figure 7; a solar chimney and Zeer-pot integrated with a house.

One problem with this concept is to integrate the chamber where the food is put to dry and at
the same time achieve an airflow in order to enhance the drying. Also the Zeer-pot’s
efficiency will most probably be more enhanced from a blowing solar dryer than a sucking
one with the same wind velocity, since the airflow will be more concentrated around the pot.
Since the solar chimney is rather large and will have to be connected to the house through

18
either a wall, a window or the roof, the flexibility will most probably be reduced depending
on how it is built.

19
 5.   Market  analysis  
The performance of the Zeer-pot is enhanced in a dry climate and mainly useful for the
population in the rural areas without access to electricity [6]. The need for cooling food is
more eminent in hot countries and in areas with no, or limited, access to electricity. The area
selected to consider for developing a market in is mapped in Figure 8 and is selected based on
the low relative air humidity and the fact that a majority of the population does not have
access to electricity.

Figure 8; area in focus [28].

Other areas that could be considered to develop a market considering their low relative
humidity is Australia, the Middle East and a few areas in Asia, for instance India. This is not a
consideration in this project as the majority of the population in the mentioned regions have
access to electricity, a more consistent food supply and storage possibilities [29].

The data in Table 1 is used to show the proportion of the population and access to electricity
in the countries in the area in question.

20
 Table 1; Population and access to electricity [4].

Percentage of the
population with
Country Population [mil.] access to Without electricity
electricity [mil.]
[%]
Tchad 13.6 6.4 12.7
Mali 17 25.6 12.7
Niger 19.1 14.4 16.3
Nigeria 177.4 55.6 78.8
Burkina Faso 17.6 13.1 15.3
Kamerun 22.7 53.7 10.5

An award, called the Rolex award, was in 2005 given to the Nigerian Mohammed Bah Abba
who designed a Zeer-pot, which can cool the air inside of it to a temperature that differ up to
14 ° from the ambient temperature. He produced and distributed about 100 000 pots to the
rural population in Nigeria [5]. Hence, the market in Nigeria is already somewhat evolved and
this project is aspiring to analyze the potential for a novel market. Consequently, Burkina
Faso is an acceptable candidate with a great amount of people living without electricity, and
since it is a country with an estimated 90 % of the population involved in the agricultural
farms or fields who could profit from enhanced storage possibilities [30].

The poorest people benefit the most from using Zeer-pots to store their food. As
approximately 46% (2009) of the population in Burkina Faso live below the poverty level, on
less than 1.25 USD a day [30], the Zeer-pots have to be extremely cheap, donated by aid
organizations or merely a precise instruction so that people can build them themselves with
whatever materials they have available.

5.1.   Burkina  Faso  

Burkina Faso is one of the poorest countries and a negligible part of the population (1% of the
women and 3% of the men) has finished secondary school. Women are an exposed group with
a gender inequality index that ranks Burkina Faso as one of the most unequal countries (131
out of 148 countries in 2012) [30]. A Zeer-pot have prospect to let the farmers sell food on
demand as the vegetables can keep fresh up to a few weeks longer. [5] This gives the women
the opportunity to sell harvested fruits and vegetables directly from their home, sparing them
from the need to walk to the market and giving them the possibility to use the time to attend
school. This is a documented effect from Mohammed Bah Abbas work in Nigeria [5].

Agriculture employs about 90 % of the population and stands for 30 % of the country’s GDP.
The most produced crops are cotton, millet, sorghum and maize [30]. The individual
consummation consists mostly of the last three mentioned, as well as rice and leaf vegetables.
A majority of people is therefore reliant on the agricultural sector not only for their income,
but also for their own food consummation [31]. Improving storage opportunities and products
lifespan by storing food in a Zeer-pot can therefore be useful.

21
Burkina Faso is an old French colony, becoming an independent state in 1960 [32]. In the
brief 50 years as an independent country it has suffered several military coups, the latest one
in 2015 [32] [33]. With this violent modern history in mind, political stability for the country
cannot be taken for granted for the nearest future. This could make it difficult for a new
business, in this case selling and distributing Zeer-pots, to be developed and reached out to
the people in need.

The quality of the school education can be questioned. Although the attendance was 60 % in
primary school in 2010, the literary rate of young people barely reached 30 % [34]. An
alternative to sell the Zeer-pot integrated with a solar dryer is to distribute a manual so that
people can build it themselves with whatever materials they have available. This is not a
realistic idea due to the high percentage of analphabets in the country, which means that the
people in the most need of this concept will not be able to benefit from it.

5.1.1.   Climate    

The country is classified into three different climate zones, illustrated in Figure 9: warm
desert climate, warm semi-arid climate and tropical savanna climate. A small area positioned
in the north, close to and influenced by the Saharan desert, has warm desert climate, and is
characterized by its constant heat, high sun duration, dry climate and windiness [35]. The heat
creates a demand for a Zeer-pot and the wind and dry climate improves its functioning.

Figure 9; Burkina Faso divided into different climate zones by the Köppen climate
classification [29].

22
The tropical savanna climate is characterized by a short four-month rainy season, with hot,
rainy weather with a lot of precipitation and a longer winter season with dryer weather and a
more comfortable temperature [35]. As a result of the high relative air humidity in the rainy
season, the Zeer-pot is not sufficient for that period of time. However, thanks to the rain
period, the land makes it suitable to grow crops.

The transition between the tropical savanna climate and the warm desert climate is called the
warm semi-arid climate. This means that the rainy period is shorter in the north of this area
[35].

5.1.2.   Business  plan  

A business profiting from selling Zeer-pots is not an option; the poor population, the people
benefiting the most from the Zeer-pots, will not be able to buy them. An alternative to this is
selling Zeer-pots extremely cheaply so that people can afford them. This is not a good option
because of the political instability in the country, which makes it hard to move freely in the
area.

Another alternative is to give people a precise instruction so they can build the Zeer-pot with
available material. A majority of the population is analphabets, as mentioned above, and this
means that the instruction will have to be oral in order to make this alternative work in
practice. This comes with some complications; as oral instructions are hard to spread
maintaining their exact content.

A third alternative and the one this study recommends, is to have the Zeer-pots donated from
help organizations. This makes it easier to reach out to all areas, as a help organization is
neutral and can move freely even in politically instable areas. This will reach out to more
people even in a politically stable country as it does not have to be limited to one or a few
locations, like an ordinary business has to. Donations require an organization that recognizes
the demand and use of the Zeer-pots and wants to sponsor them; this may take some time to
establish. The clay material could be produced directly in the country as the raw material
exists [36].

In 1999 Mohammed Bah Abba sold and distributed Zeer-pots in Nigeria for 2 US dollars for
the small pots and for 4 US dollars for the bigger ones [5]. Distributed in large quantities the
price could potentially be reduced even more. To produce one Zeer-pot only two clay pots
and a frequent refill of sand, which many people have access to in the chosen area, is
required. One or two Zeer-pots is enough to cool one family’s food for many years, as the
clay pots have a long lifespan when handled with reasonable care. This makes it an
investment worth considering for help organizations.

23
 6.   Methodology  
To attain a correlation for how many degrees more the temperature drops and how much
faster a final temperature is reached for increasing wind velocities, tests were made on the
same pot in otherwise as constant conditions as was practically possible.

Two larger and two smaller pots, fitting into the larger ones, were purchased. One Zeer-pot
was then tested only for wind velocities and the other only for the case of the pot hanging in
the air and also for being glazed on the inside.

A reference case was designed and tested in order to compare the pot glazed on the inside and
the pot hanging in the air in the same conditions as the reference test.

An optimal experiment setup would have been to run each test with the same relative
humidity, but this was not practically possible in the given lab conditions. A higher relative
humidity for a test results in a higher wet bulb temperature, the lowest theoretical
temperature, and the values for the final temperatures is not directly comparable between tests
with different relative humidity. Therefore, a way to compare these results was to calculate
the final temperature the pot achieved in relation to the lowest possible theoretical
temperature, the wet bulb temperature.

A formula was created to make the results comparable:

Lx NLyz{ml|{
, (10)
Lx NL}{n  ~•€•

where 𝑇X is the ambient temperature, 𝑇cUjbc‚j is the average temperature between the lower
temperature and the upper temperature in the inner pot and 𝑇ƒji„e…„ is the wet bulb
temperature. This allowed a comparison independent of the relative humidity by calculating a
percentage of the temperature drop between the wet bulb temperature and the final
temperature.

6.1.   Experiments  and  their  setup  

The experiment was arranged in a climate chamber, which had an approximately constant
ambient temperature of 40  ℃. A dehumidifier which was constantly on maximal effect and an
adjustable fan was installed to attain different wind velocities in front of the pot. The pots
were placed on a wooden plate about 10-15 cm from the fan, depending on the wind velocity,
see Figure 10.

Temperature measurements were logged and four different temperatures was always
measured in each experiment:
•   The ambient temperature
•   The sand temperature
•   The air temperature on the inside of the upper part of the pot
•   The air temperature on the inside of the lower part of the pot.

24
 Figure 10; the lab setup with the dehumidifier to the left and the pot with the fan
behind it to the right.
Humidity measurements were made continuously and an average value, the value which was
held most continuously through the test, was determined from these measurements. This is
seen in table 2. How much the relative humidity varied for each test can be seen in Table 7 in
Appendix 1.

The measurements were started when the sand temperature reached 38  ℃ and was continued
until the temperature decrease was negligible. The time to reach their final air temperature in
the pot as well as the value of the final temperature was compared between the tests.

The climate chamber was frequently opened in order to regularly refill the pot with water
(same as sand temperature), to re-soak the towel and to empty the dehumidifier. This resulted
in temporary dips of the inside temperature, seen in the results graph.

25
 7.   Results    
In this section the results from the experimental studies are presented.

7.1.   Results  for  the  wind  velocity  tests  

One pot, named pot 1 during the experiments, has been exposed to different wind velocities
and the relative humidity in the air, the time to a final temperature and the value of the final
temperature have been measured. All of these values from the experiments are presented in
Table 2. The graph for each test is presented in Appendix 2.

Table 2; final temperature and cooling time depending on wind velocity.

Time to final
Final upper Final lower temperature Average
Wind velocity temperature temperature from a sand relative
[m/s] [°𝐶] [°𝐶] temperate of humidity
38°𝐶 [%]
[minutes]
0.1-0.2 30 28.8 230 38
0.3-1 31 28.5 160 41
1-2 31.3 30.9 260 40
2-2.5 31.6 31.6 165 49
3-3.5 31.8 31.5 135 53
5-6 32.6 32.6 80 55
6-7 33.3 29.2 50 43

How the time to reach the final temperature is influenced by the different wind velocities is
presented in Figure 11. A relatively linear correlation can be seen in this graph, apart from the
second and third value from the left.

26
 Figure 11; the different times to reach the final temperature depending on the wind
velocity.

In order to compare the tests with the differences in relative humidity, the wet bulb
temperatures is derived from a Mollier diagram, see Appendix 3, and presented in Table 3.
The percentages calculated with Equation 10 of the relation between the final temperature and
the wet bulb temperature is also presented in Table 3. The weighted temperature decrease is
presented in Figure 12.

Table 3; relevant data compared for the cases of different wind velocities.

The final
Final average Average Average Wet bulb temperature
Wind velocity temperature relative ambient temperature in
[m/s] between lower humidity temperature [°𝐶] percentage
and upper [%] [°C] compared to
temperature the wet bulb
[°𝐶] temperature
[%]
0.1-0.2 29.4 38 40 27 82
0.3-1 29.8 41 40 28 85
1-2 31.1 40 40.5 27.5 72
2-2.5 31.6 49 39 29 74
3-3.5 31.7 53 39 30 81
5-6 32.6 55 39.5 31 82
6-7 31.2 43 40.5 29 81

27
Figure 12; the temperature decrease in percentage of the lowest theoretical temperature
for different wind velocities.

28
 7.2.   Results  for  the  hanging  construction  
Two tests were set up with pot 2 in order to examine whether a Zeer-pot standing on the
ground is more efficient than a hanging construction. The result for the first test, which is the
reference scenario for both the hanging and the glazed Zeer-pot, is presented in Figure 13.
Collected data for the reference test is found in Table 4. The result for the second scenario,
the hanging construction, is presented in Figure 14. Relevant data for this test is found in
Table 5. These tests are evaluated in the same way as the wind velocity tests, by how much
the final temperature decreased in relation to the wet bulb temperature.

Figure 13; a graph of the results for the reference case.

Table 4; data for the reference case.

The final
temperature in
Wind Ambient Final Average Wet bulb percentage
velocity in temperature temperature humidity temperature compared to the
front of the [°𝐶] [°𝐶] [%] [°𝐶] wet bulb
pot [m/s] temperature [%]
3-3.5 38.7 32 38 26.5 55

29
 Figure 14; a graph for the case of the pot hanging in the air.

Table 5; data for the case of the Zeer-pot hanging in the air.

The final
Wind temperature in
velocity in Ambient Final Average Wet bulb percentage
front of the temperature temperature humidity temperature compared to the
pot [m/s] [°𝐶] [°𝐶] [%] [°𝐶] wet bulb
temperature [%]
3-3.5 39.7 33.2 48 29.5 64

30
 7.3.   Results  for  the  glazed  pot  
One more test was set up to examine how much glazing on the inside decreases the cooling
capacity. These results can be seen in Figure 15 and Table 6. The result for the reference case
is the same as for the previous section and is presented in Figure 13 and Table 4.

Figure 15; a graph for the case of the pot glazed on the inside.

Table 6; data for the case of the glazed Zeer-pot.

The final
Wind temperature in
velocity in Ambient Final Average Wet bulb percentage
front of the temperature temperature humidity temperature compared to the
pot [m/s] [°𝐶] [°𝐶] [%] [°𝐶] wet bulb
temperature [%]
3-3.5 40.1 32.2 41 28 65

31
 8.   Discussion  and  conclusions  
An almost linear correlation between wind velocity and the time to reach the final temperature
can be observed, see Figure 11. This apart from two values that differ from this trend, due to
incorrectly performed experiments, see the section 8.2. Specific sources of error.

The course of the temperature decrease in relation to the wet bulb temperatures is seen in
Figure 12. A nicely shaped curve can be observed, apart from the two first values that differs
from this trend, see the section 8.2. Specific sources of error. A slight stagnation in the
increase is noted for the last values of air velocity, which is reasonable since the wind velocity
cannot increase the cooling capacity to infinity. From this curve, it can also be interpreted that
the wind velocities between 3-5.5 m/s has the greatest effect on the Zeer-pots efficiency,
although the results for lower wind velocities is not really reliable due to insufficient
information from the experiments. Thus, if a solar dryer is combined with the Zeer-pot, more
tests would have to be done in order to be able to make some concrete conclusions regarding
the wind velocity that is needed to make a significant difference. But a value between those
mentioned above is guaranteed to make the Zeer-pot more efficient.

In the comparison study of the hanging construction and the pot standing on the ground, see
7.2. Results for the hanging construction, the hanging pot reached a 64 % decrease relative to
the wet bulb temperature in comparison to 55 %, which was the result for the reference study.
The hypothesis was that for the hanging constructions an additional exposure to airflow on the
underside of the pot would enhance the cooling capacity. This was confirmed by the test
where an increase in the cooling capacity was noticed.

The glazed pot reached a 65 % decrease in relation to the wet bulb temperature, in
comparison to 55 % for the reference study. According to these results, glazing the inside of
the pot in this environment, in order to prevent the food on the inside for exposure to
moisture, does improve the cooling capacity. This is a strange and invalid result, since the
glazing is supposed to prevent the heat transfer and the evaporation. See more about why this
test didn’t show the expected result in 8.2. Specific sources of error.

The simulated Zeer-pot described in 3.7. Simulations of the Zeer-pot reached a decrease of 90
% relative to the wet bulb temperature which differed a lot from the experimental results. For
pot 1, where the wind velocity was 3-3.5 m/s, a decrease of 81 % was accomplished and for
the reference case for pot 2 the decrease was 55 %. The reason the experimental values are
lower than the simulated value is probably mainly due to the amount of radiation in the
chamber, and the fact that pot 2 had even worse results is probably due to that it did not let as
much water move through the clay. Both of these factors are explained more in 8.1. General
sources of error.

For practical reasons a larger Zeer-pot could be described as more fitted to store larger
amounts of food in than a few small pots. The cooling will be more efficient though with a
large surface area compared to the volume, as the cooling capacity depends on the ratio
between the surface area and volume [20]. Think about it as if an area is supposed to provide
heat to a volume of air. If the volume is big it will take a lot more energy to heat than if it is a
small volume. As the size of the Zeer-pot becomes larger, the volume increases to the cube of

32
the linear dimensions and the surface area increases to the square of the linear dimensions.
Therefor a large volume results in a less effective system. This has been taken into account
when the size of the pots was chosen (29 cm in diameter and 24 cm in diameter) in order to
provide room for some content and at the same time minimize the size.

8.1.   General  sources  of  error  

One of the reasons that the results were not fully in accordance with the theoretical values,
could be due to the fact that equilibrium in the heat transfer through the pots were not
reached. This means that the heat is not only drawn out from the pots due to temperature and
humidity differences, but that heat is also transferred to the pots from its surroundings. The
fact that the climate chamber used during the experiments was rather small, a cube with a side
roughly around 1.5 meter, and dressed with metal on the inside, results in radiation from the
surrounding walls affecting the pots. The dehumidifier, which had to be placed rather close to
the pots, radiated heat as well and therefore the pots are constantly exposed to radiated heat
from the surroundings. This affects the Zeer-pot in a way that prevents the inside temperature
from reaching a temperature as low as the wet bulb temperature. In a real life scenario this
much radiation may not be the case, since the place where the pots will be stored may have a
larger space around them, walls that are not made of metal maintaining a temperature of 40°C
and there will not be a warm dehumidifier standing right next to it either. If this is the case,
the Zeer-pots efficiency will increase and a lower temperature is achievable.

Two other problems caused by the experimental setup was that the relative humidity could not
be adjusted and that the airflow could not be as controlled as desired, nor could it be made
laminar. Since the relative humidity changed both during and between experiments it was
hard to make any direct comparison of the results. Running the experiments in the climate
chamber without the dehumidifier to keep it even was not an option either, since the air was
not replaced rapidly enough to prevent the relative humidity from spiking to 100 %. Thereof
the comparison chosen to be made was depending on how close the final temperature was to
the wet bulb temperature measured in percentages, for an average relative humidity measured
in every experiment. This could be considered a source of error and the results would be a lot
easier to interpret if the relative humidity was the same in each test. The fan put in the climate
chamber to create wind was not able to create a laminar flow, which resulted in large
variances in the measured wind velocities for every test. A mean value had to be chosen,
which is not necessarily completely accurate. Although, the airflow will mostly be turbulent
when the pots are used in real life, at least if they are placed outdoors, which means that the
tests are relevant from that point of view.

In most of the graphs a larger variation in temperature can be observed, especially in the
ambient air. This is because the climate chamber had to be opened while, for example, adding
water to the sand or changing a dry towel to a new wet one, and therefore a disturbance was
created in the system. Although the disturbance usually did not continue for more than maybe
one minute, it could still have impact on the final result. If the Zeer-pot would be used or
tested in either a bigger climate chamber or in the environment, in which it is meant to be
used, this would not have been an issue since the presence of an opened door would not affect
the system to the same extent.

The amount of moisture in the sand was another problem during the experiments. The
intention was to always have the same amount of water in the sand in the beginning of the

33
experiments, but the time available to complete the experiments was limited. This caused the
need to start a second experiment almost right after the first one was finished, and as a result
the sand was almost always at a somewhat moist state. It was also hard to determine how
moist it was since only the top layer was visible to inspect. An approximation then had to be
made to have a tolerably even amount of water in the experiments.

To make the pots temperature keep decreasing as the water evaporates and the sand is drying
out, water had to be refilled regularly. Trying to avoid disturbing the systems rhythm, the
water added was always kept at a temperature close to the one in the sand. Although in real
life applications the opportunity to adjust the temperature of the water will generally not be an
option, which will probably make the Zeer-pot take longer time to reach the final temperature
which in turn may be a little bit higher.

As the sand kept getting moisturized with water throughout the experiments it was observable
that the top layer of the sand kept moving downwards between the refills. This is probably
due to that the sand was dry when placed between the pots. As the water is added it fills the
spaces between the grains, making them a little bit bigger. Then as the water evaporates,
leaving the spaces empty, the sand will get packed tighter than before adding the water,
causing a decreased sand volume.

Another problem during the experiments was that the hygrometer used was not showing
accurate measurements. When compared to an old-fashioned mechanic hygrometer showing
the dry- and wet bulb temperatures, through which the relative humidity was found with
Molliers diagram, it turned out that the one used in the climate chamber was about 16% off.
This affects the percentages of temperature decrease calculated in 7. Results. The wet bulb
temperature does not vary exactly linearly with the relative humidity in the Moillier diagram,
see Appendix 3, and consequently each experiment will be affected differently. Despite this,
the difference cannot be considered to be more than maybe one or two degrees, and this can
be neglected on account of the accuracy of the experiments.

Upon purchase, the pots were chosen based mainly on three qualities; available sizes, that
they were made out of clay and that they were unglazed. The pots that were bought had all of
those three qualities. However, they had earlier been processed by being heated to a rather
high temperature while drying, causing the surface to harden. This resulted in the pot not
letting the water move through the clay as easily and thereof aggravating the evaporation and
cooling of the pots. To facilitate the evaporation, the pots were polished with sandpaper to
remove the surface. While polishing, the naked eye could roughly estimate that the surface
did not become perfectly even. Consequently, the pots may not evaporate exactly the same
amount of water everywhere, even though the differences will probably be small. Depending
on which side chosen to be facing the fan, this may have had impact on the final results.

The original intention was to keep the pots at around 38°C throughout both sand and clay
before starting the experiments. The only way to do this was to keep the pots inside the
climate chamber for a few hours before each test. Opening the door to the climate-chamber
allowed cooler air to come in from the outside room, keeping it open for a longer amount of
time would lower the temperature of both the air and the pots, which would affect the results
of the experiments. Therefor it was not possible to wait and see how much the water would
sink into the sand.

34
 8.2.   Specific  sources  of  error  
The percentage of the final temperature relative to the wet bulb temperature, seen in Figure
12, shows a rather nice shape of the curve between approximately 1.5-5.5 m/s. The fact that
the tests with the two lowest velocities stand out significantly does not seem like a reasonable
result.

For the test with a velocity of 0.3-1 m/s the towel was changed and the sand was watered at
the time that the dip is shown in the graph, see Appendix 2, around 100 minutes after the test
was started. Since the towel was changed at that moment there is a rather large possibility that
cold air accidently entered from outside the climate chamber got into the inner pot. Maybe the
water used was also on the verge of being too cold. After the finished maintenance of the pot,
as it returned to a state without outside influence, the temperature inside was already lowered.
This means that as it continued to evaporate and cool, the temperature was lowered even more
due to the cold inside. Maybe the walls in the climate chamber had the time to cool down and
did not radiate as much heat for a while after opening, meaning that a value closer to
equilibrium in the heat transfer is achievable and thereof a lower final temperature. The fact
that mainly the lower inside temperature decreases, the most reasonable conclusion is that
some sort of disturbance influenced it. No obvious reason was found after the test was
finished and repeating the test to compare the results may solve the problem.

The experiment regarding pot 1 with a wind velocity of 0.3-1 m/s, see Appendix 2, was never
able to be finished completely, since the lab was closing for the day and there was no time to
run it again. The data and correlations from the different experiments indicates that it would
have taken around 210 minutes, which seems like a reasonable amount of time. The dip that is
shown in the graph is due to refilling the sand with water. This can also be observed in the
experiment for a velocity of 1-2 m/s, see Appendix 2, where the temperature curves for the
pots start moving upwards at 120 minutes. This means that the sand between the pots has
started to dry out due to the evaporation and can no longer provide the same heat transfer and
cooling. Unfortunately, this was not discovered in time to refill the water before the
temperature increased, probably due to the fact that the amount of water in the sand was not
sufficient, as explained above.

Unfortunately no obvious interruptions of the test for 0.1-0.2 m/s can be seen in the graph, see
Appendix 2, yet the test result is much better than expected. One reason why this test turned
out to be one of the best may be that this was one of the last tests performed. The methods
were refined by then and better routines were established as how to execute the experiments
without complications. However, this cannot be considered the only reason, since there is a
major difference compared to the test of 1-2 m/s for instance.

Regarding the test for pot 2 with a glazed inner pot, the results achieved for this test was not
the same as expected. The inner side of the inner pot was painted with pot paint which did not
really stick to the surface; the pot was painted several times due to this. Once the experiment
was performed and the Zeer-pot was brought out of the climate chamber, the paint had started
to crack and also peel on the top, closest to the towel. The cracks and the areas starting to peel
were not very big, so supposedly they did not affect the test results too much, but it is not
possible to exclude this as a source of error. One possibility explaining the results would be
that the pot paint blocks the radiation coming from the surroundings, insulating the inner pot

35
and helping it achieve a lower final temperature. Otherwise there are no apparent reasons in
the experimental setup that would point to these results. A repeated test is recommended.

While comparing results for the same wind velocities of 3-3.5 m/s for pot 1 and pot 2, see
Appendix 2 and Table 3 for pot 1 and Figure 13 and Table 4 for pot 2, an 81% decrease in the
final temperature relative to the wet bulb temperature for pot 1 and only a 55% decrease for
pot 2 can be observed. This means that the pots are not evaporating equally much due to
polishing with sandpaper, see 8.1. General sources of error. It could also be noticed that the
pots were not equally hardened, since pot 2 did not let as much water pass through its walls
from the beginning, even before we polished them. Although they were pretty much equally
polished, the differences are apparently rather large anyway. Due to uncertainties in their
different qualities before the start of the experiments, these results are a confirmation of that
splitting up the tests for wind velocity and hanging/glazed between the two pots was a good
decision.

Since the experiments were only performed once for every velocity and individual setup, the
results cannot be analyzed statistically. To create a more extensive study it needs to be built
up with more tests on the same velocities, hopefully proving the results and conclusions in
this report to be accurate, and of course, statistically significant.

8.3.   Conclusions  
Apart from two values that differ, an almost linear correlation between the time to reach the
final temperature and wind velocity was observed, even though a stagnation in the end of the
interval is observed.

The temperature decrease in relation to the wet bulb temperatures is linear in the beginning,
except from two values that differ, with a slight stagnation at the end.

The hypothesis was that for the hanging constructions an additional exposure to the airflow on
the underside of the pot would enhance the cooling capacity. This was experimentally
confirmed in one test, as an increase in the cooling capacity was observed.

No unequivocal conclusion can be drawn for the tests regarding the pot glazed on the inside.
Further experiments have to be done in order to reach a conclusion on how glazing affects the
cooling capacity.

If a solar dryer is combined with the Zeer-pot, an airflow with a wind velocity around 3-3.5
m/s is guaranteed to improve its cooling capacity, but more tests on lower velocities needs to
be done to find a minimum.

36
 8.4.   Recommended  future  work  
Although the phenomena with evaporative cooling is relatively well-studied, the specific
Zeer-pot application is not as covered. There are a lot of factors and parameters that could be
examined to optimize the cooling capacity. This could make it easier to establish the concept
on the market.

If this study would have had more time to run tests on the pots the different wind velocities
would be tested again to verify the results and make it statistically valuable. Smaller intervals
of the different velocities would be tested too to get a more precise curve of results. Also the
hanging construction and glazed inner pot would be tested again to see if the incoherent
results are correct or should be discarded.

Tests that would have been done if there would have been more advanced equipment
available or the accurate climate is to test for a fairly constant relative air humidity to get
results that are more accurate and easier to compare.

Other parameters that could be examined is:

•   Whether a clay cap or a wet towel is more efficient.
•   The optimal size of the Zeer-pot for both efficiency and practical aspects.
•   The best distance for the sand gap.
•   The optimal clay and sand density.
•   The most optimal materials based on efficiency, price and accessibility.

Applying the solar dryer and making calculations/tests on it to see what wind velocities is
possible to create and also investigate further to see which type of solar dryer should be used
is also an important part for the development of this concept.

37
 9.   References  

[1]
https://www.wfp.org/hunger/stats
World food program, WFP
Accessed: 2016-04-24

[2]
http://www.fao.org/docrep/014/mb060e/mb060e02.pdf).
“Extent of food losses and waste”
Food and Agriculture Organization of the united nations, FAO
Accessed: 2016-04-24

[3]
http://www.fao.org/docrep/014/mb060e/mb060e.pdf
“Global food loss and waste”
Food and Agriculture Organization of the united nations, FAO
Accessed: 2016-04-25

[4]
http://data.worldbank.org/indicator/EG.ELC.ACCS.ZS
The world data bank
Accessed: 2016-03-18

[5]
Rolex awards, Mohammed Bah Abba
http://www.rolexawards.com/profiles/laureates/mohammed_bah_abba/project
Published: 2005
Accessed: 2016-03-20

[6]
ZOLARMAD Zeer pot refrigerator- Heat and mass transfer analysis
Sebastian Haglund El Gaidi
28/7 2014 Majunga, Madagaskar

[7]
Picture.
http://survivestayalive.com/keep-food-cool-without-electricity-zeer-pots
Accessed: 2016-03-15

[8]
Practical Action and the Woman’s Association for Earthenware Manufacturing
Sudan
http://www.fao.org/climatechange/17850-0c63507f250b5a65147b7364492c4144d.pdf
Accessed: 2016-05-12

38
[9]
Postharvest Technology and Food Process Engineering
Amalendu Chakraverty, R. Paul Singh
Accessed: 2016-05-16

[10]
A Review of Evaporative Cooling Technologies
O. Amer, R. Boukhanouf, and H. G. Ibrahim
http://www.ijesd.org/vol6/571-W4004.pdf
Accessed: 2016-05-08

[11]
“Effect of the Evaporative Cooling Techniques by Spraying Mist Water on Reducing Urban
Heat Flux and Saving Energy in Apartment House”
Daisuke Narumi, Kentaro Shigematsu and Yoshiyuki Shimoda,
Osaka University Suita City, Osaka Pref., Japan
http://heatisland2009.lbl.gov/docs/221000-narumi-doc.pdf
Accessed: 2016-05-06

[12]
Çengel, Yunus A., Ghajar, Afshin J. 2015. Heat and mass transfer - Fundamental &
Applications. Fifth edition. New York: McGraw-Hill Education. Pages 17-18.

[13]
Çengel, Yunus A., Ghajar, Afshin J. 2015. Heat and mass transfer - Fundamental &
Applications. Fifth edition. New York: McGraw-Hill Education. Pages 379-388.

[14]
Picture
https://flutuante.files.wordpress.com/2013/03/pot_in_pot.jpg
Accessed: 2016-05-05

[15]
Çengel, Yunus A., Ghajar, Afshin J. 2015. Heat and mass transfer - Fundamental &
Applications. Fifth edition. New York: McGraw-Hill Education. Table A-15 page 924.

[16]
Çengel, Yunus A., Ghajar, Afshin J. 2015. Heat and mass transfer - Fundamental &
Applications. Fifth edition. New York: McGraw-Hill. Pages 28-29.

[17]
Ekroth, Ingvar, Granryd, Erik. 2013. Tillämpad Termodynamik. Edition 1:4. Lund:
Studentlitteratur AB. Pages 520-524.

[18]
http://www.engineeringtoolbox.com/dry-wet-bulb-dew-point-air-d_682.html
Dry bulb, wet bulb and dew point temperatures
The engineering toolbox
Accessed: 2016-03-10

39
[19]
Ekroth, Ingvar, Granryd, Erik. 2013. Tillämpad Termodynamik. Edition 1:4. Lund:
Studentlitteratur AB. Page 525.

[20]
http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html
Thermal Conductivity of some common Materials and Gases
Accessed: 2016-04-20

[21]http://www.academia.edu/16500805/Experimental_Study_of_Alternatives_to_Sand_in_Z
eer_Pot_Refrigeration_Technique
Experimental  Study  of  Alternatives  to  Sand  in  Zeer  Pot  Refrigeration  Technique
Prabodh Sai Dutt R and  Thamme  Gowda  C.S. at Vidyavardhaka College of Engineering, India.
Vol. 5, Iss. 5, May 2015
Accessed: 2016-04-29

[22]
https://knoema.com/UNSDESD2015/un-statistics-division-energy-statistics-database-
2015?location=1000320-burkina-faso
UN  Statistics  Division  Energy Statistics Database, 2015
Accessed: 2016-04-29

[23]
https://energypedia.info/wiki/Burkina_Faso_Energy_Situation
Burkina Faso Energy Situation
Accessed: 2016-05-03

[24]
Jaconelli, Livio, Jaconelli Palm, Kim. 2016. A thermal network model for a combined
evaporative cooling system with a solar chimney. Stockholm: The Royal Institute of
Technology. (Bachelor of Science Thesis for the School of Industrial Engineering and
Management. Energy Technology EGI-2016)

[25]
http://www.sciencedirect.com.focus.lib.kth.se/science/article/pii/S1364032108001251
Solar energy driving systems– a review
Aful Sharma, C.R Chen, Nguyen Vu Lan
Published: 2008
Accessed: 2016-03-15

[26]
http://www.sciencedirect.com.focus.lib.kth.se/science/article/pii/S1364032109001567
Review of solar dryers for agricultural and marine products
A. Fudholi, K. Sopian, M.H Ruslan, M.A Alghoul, M.Y Sulaiman
Published: 2009
Accessed: 2016-03-19

40
[27]
http://www.sciencedirect.com.focus.lib.kth.se/science/article/pii/S0196890405002554
Multi shelf domestic solar dryer
Parm Pal Singh, Sukhmeet Singh, S.S Dhaliwal
Published: 2005
Accessed: 2016-03-16

[28]
http://www.intellicast.com/Global/Humidity.aspx?region=wafri
Relative Humidity
Accessed: 2016-03-25

[29]
http://www.intellicast.com/Global/Humidity.aspx
Global humidity map
Accessed: 2016-03-18

[30]
Country fact sheet on food and agriculture policy trends, April 2014
“Socio-economic context and role of agriculture”
Food and Agriculture Organization of the united nations, FAO
http://www.fao.org/docrep/field/009/i3760e/i3760e.pdf
Accessed: 2016-03-29

[31]
Monitoring African Food and Agricultural Policies Project (MAFAP),
Commodity selection in the five pilot countries
http://www.fao.org/fileadmin/templates/mafap/documents/Commodity_Selection_MAFAP_0
8-10.pdf
Accessed: 2016-05-18

[32]
“Burkina Faso”
James Fearon David Laitin
Stanford University, July 2006
http://web.stanford.edu/group/ethnic/Random%20Narratives/BurkinaRN2.4.pdf
Accessed: 2016-05-17

[33]
International business times, September 2015
http://www.ibtimes.com/burkina-faso-elections-2015-coup-transitional-government-
jeopardizing-fragile-2106834
Accessed: 2016-05-17

[34]
The World bank, January 2016
http://www.worldbank.org/en/country/burkinafaso/overview
Accessed: 2016-05-18

41
[35]
Updated world map of the Koppen-Geiger climate classification
M. C. Peel, B. L. Finlayson and T. A. McMahon
Department of Civil and Environmental Engineering,
The University of Melbourne, Victoria, Australia
http://www.hydrol-earth-syst-sci.net/11/1633/2007/hess-11-1633-2007.pdf
Accessed: 2016-05-16

[36]
Mineralogy of clay raw material from Burkina Faso and Niger used for ceramic wares
Tibo Siméon Kabre, , Karfa Traore, Philippe Blanchart
http://www.sciencedirect.com/science/article/pii/S0169131798000052
Accessed: 2016-05-15

42
 10.  Appendix  
Appendix 1; Variations in relative humidity during the tests

Table 7; how the relative humidity varied in each test.

Pot 1 Relative air humidity [%]

0.1-0.2 m/s 31-41
0.3-1 m/s 34-45
1-2 m/s 35-41
2-2.5 m/s 42-51
3-3.5 m/s 44-59
5-6 m/s 50-56
6- m/s 41-46
Pot 2
Reference case 32-39
Hanging construction 40-48
Glazed inner pot 38-44

43
Appendix 2; Graphs for wind velocity tests

44
45
Comment: The sand temperature is lower than in the other graphs most probably due to air
leaking through the towel.

46
47
Appendix 3; Used Mollier diagram

48