Design and Development of a Field Monitoring Control System for a Green Roof

Bachelor of the Science of Engineering

Final Year Project

Department of Electrical Engineering

University of Moratuwa

DESIGN AND DEVELOPMENT

OF A FIELD MONITORING

CONTROL SYSTEM

FOR A GREEN ROOF

Supervised by Dr. A. G. Buddhika Jayasekara

Dr. D.P. Chandima

B.M.A.N. Balasooriya 090040D 2-1

A.G.N. Bandara 090041G 2-2

H.G.I.W. Bandara 090044T 2-3

K.S. Buddhasiri 090057K 2-4

ii | P a g e

PREFACE

This report represents the steps of design and development of a field monitoring control

system for a green roof which is for the final year project in partial fulfillment for the

requirement of the Bachelor of the Science of Engineering in the Department of Electrical

Engineering, University of Moratuwa.

In chapter 1, the introduction to green roofs and the need of irrigation for a green roof and are

mentioned. Currently used irrigation methods and mathematical models which are already

available, to be used within our project are described in the chapter 2. A study for justifying

the need of optimized irrigation for green roofs in different zones in Sri Lanka is described in

the chapter 3. Control algorithm to implement the irrigation scheme is mentioned in the chapter

4, while chapter 5, 6 and 7 shows how electrical design was done, how monitoring interface

was created and hoe mechanical structure was constructed, respectively. Finally a summary

about the project is mentioned under conclusion and additional documents are attached in

annexes & appendices.

Here we have tried our best to include all the necessary information about the project.

iii | P a g e

ACKNOWLEDGMENT

First and foremost we would like to express our sincere gratitude to our project supervisor

Dr. Buddhika Jayasekara and Dr D.P Chandima for the immense support throughout the year

to guide us in the correct path at all times to achieve the targets set to us.We would like to

extend our thanks to the final year project coordinator Dr.Harsha Abeykooon. Our deep sense

of appreciation goes to the academic staff of the department of Electrical Engineering,

University of Moratuwa, for the invaluable advices and guidance. A special thank you should

be noted here for all the assistance rendered to us by the non-academic staff of the

department of Electrical Engineering, University of Moratuwa. Lastly we thank all the

individuals who assisted us in making this endeavour a success.

iv | P a g e

DECLARATION

The work submitted in this project report is the result of our own combined investigation,

except where otherwise stated. It has not already been accepted for any degree, and is also not

being concurrently submitted for any other degree.

………………………………. ………………………………….

B.M.A.N. Balasooriya (090040D) A.G.N. Bandara(090041G)

………………………………. ………………………………...

H.G.I.W. Bandara(090044T) K.S. Buddhasiri(090057K)

I endorse the declaration by the candidates.

…………………………. …………………………….

Dr. A.G.B.P. Jayasekara Dr. D.P. Chandima

v | P a g e

ABSTRACT

Historically, green roofs were designed for natural precipitation with a plant selection

focusing on hardy succulents that can survive harsh, water stressed conditions. Although this

seems a convenient solution to establish and maintain a green roof system, at a much broader

level this does not optimize the functions and performance of the green roof. In this report the

need for optimization of irrigation for a green roof and design and development of an irrigation

control system based on water conservation and weather adaptability are discussed. An

irrigation control system is designed base on both volumetric water content and

evapotranspiration and implemented in a laboratory prototype of a green roof.

vi | P a g e

Table of Contents PREFACE .................................................................................................................................. ii

ACKNOWLEDGMENT.......................................................................................................... iii

DECLARATION ...................................................................................................................... iv

ABSTRACT ............................................................................................................................... v

LIST OF FIGURES ............................................................................................................... viii

LIST OF TABLES .................................................................................................................... ix

1. INTRODUCTION .............................................................................................................. 1

1.1. Green Roofs................................................................................................................. 1

1.2. Green Roof Types ....................................................................................................... 1

1.3. Green Roof Benefits .................................................................................................... 2

1.4. The Need of irrigation for a green roof ....................................................................... 2

2. LITERATURE REVIEW ................................................................................................... 5

2.1. Penman-Monteith Equation for Estimating ET ........................................................... 6

2.2. Irrigation scheduling methods ................................................................................... 12

2.3. Sprinkler placement for uniformly distribution of water .......................................... 14

3. STUDY ON THE NEED FOR OPTIMIZED IRRIGATION FOR GREEN ROOFS IN

DIFFERENT ZONES IN SRI LANKA ................................................................................... 17

3.1. Overview ................................................................................................................... 17

3.2. Simulation ................................................................................................................. 17

3.3. Amount of water loss due to ET................................................................................ 21

3.4. Conclusion of the Study for the justification of the artificial irrigation requirement.

27

4. DEVELOP A CONTROL ALGORITHM TO IMPLEMENT THE IRRIGATION

SCHEME ................................................................................................................................. 28

4.1. Overview ................................................................................................................... 28

4.2. Irrigation control by predicting the evapotranspitaion for the current hour. ............. 29

4.3. Programming ............................................................................................................. 34

5. ELECTRICAL DESIGN .................................................................................................. 36

5.1. Overall system ........................................................................................................... 36

5.2. Sensors that are used in local controller .................................................................... 36

5.3. Irrigation controlling circuit ...................................................................................... 40

5.4. Irrigation actuation .................................................................................................... 42

5.5. Wireless communication between the local controller and PC ................................. 44

vii | P a g e

6. MONITORING INTERFACE ......................................................................................... 46

6.1. Overview ................................................................................................................... 46

6.2. Data logging capability ............................................................................................. 47

7. MECHANICAL STRUCTURE ....................................................................................... 48

7.1. Design Requirements ................................................................................................ 48

7.2. Design........................................................................................................................ 48

7.3. Construction .............................................................................................................. 51

8. CONCLUSION ................................................................................................................ 55

9. REFERENCES ................................................................................................................... x

10. Annexes and Appendices ............................................................................................... xi

Annex 1.Arduino Code ......................................................................................................... xi

Annex 2.ET Library ......................................................................................................... xviii

Annex 3. NI LabVIEW Block Diagram ............................................................................... xx

Appendix 1. Logged Data ................................................................................................... xxi

viii | P a g e

LIST OF FIGURES

Figure 1-1 Green Roof Layers ................................................................................................... 2

Figure 2-1: Soil Moisture Conditions ........................................................................................ 5

Figure 2-2:Types of sprinkler Spacing .................................................................................... 15

Figure 2-3:triangular Spacing .................................................................................................. 15

Figure 2-4:Square Spacing ....................................................................................................... 15

Figure 3-1 Green Roof simulation Method .............................................................................. 17

Figure 3-2:Daily ET of Colombo (2012) ................................................................................. 19

Figure 3-3:Daily ET of Anutadhapura (2012) ......................................................................... 20

Figure 3-4:Daily ET of Hambantota (2012) ............................................................................ 20

Figure 3-5:Daily Water Balance (Colombo)............................................................................ 25

Figure 3-6:Daily Wataer Balance (Kurunegala) ...................................................................... 25

Figure 3-7:Daily Water Balance (Kurunegala) ........................................................................ 26

Figure 3-8:Daily Water Balance (Hambantota) ....................................................................... 26

Figure 4-1:Evapotranspiration Overview ................................................................................ 28

Figure 4-2: Overall algorithm .................................................................................................. 30

Figure 4-3:Modified Seasonal Decomposition method ........................................................... 31

Figure 4-4: Seasonal Decomposition for Temperature ............................................................ 32

Figure 4-5:Seasonal Decomposition for Humidity .................................................................. 32

Figure 4-6:Seasonal Decomposition for solar radiation .......................................................... 33

Figure 4-7:Moving Average for Wind Speed .......................................................................... 33

Figure 4-8:Libraries Used ........................................................................................................ 34

Figure 4-9:Connectivity of Function ....................................................................................... 34

Figure 5-2:Stephenson Sheild .................................................................................................. 36

Figure 5-1:DHT11 sensor ........................................................................................................ 36

Figure 5-3:Amplifier For Anemometer ................................................................................... 37

Figure 5-4: Anemometer Calibration Results .......................................................................... 38

Figure 5-5:Arduino Board........................................................................................................ 40

Figure 5-6:Power Supply Unit Circuit Diagram ...................................................................... 41

Figure 5-7: Power Supply Unit Hardware ............................................................................... 41

Figure 5-8: Solenoid Valve ...................................................................................................... 42

Figure 5-9:Driver Circuit for Solenoid valve........................................................................... 42

Figure 5-10:flow Sensor .......................................................................................................... 43

Figure 5-11:Flow Sensor Caliberation Results ........................................................................ 44

Figure 5-12:Xbee ..................................................................................................................... 44

Figure 5-13:Xbee Shield .......................................................... Error! Bookmark not defined.

Figure 6-1: Monitoring Interface ............................................................................................. 46

Figure 7-1: Mechanical Structure Design_Overall View ........................................................ 48

Figure 7-2: Mechanical Structure_Isometric View .................................................................. 49

Figure 7-3: Mechanical Structure_Plan View .......................................................................... 49

Figure 7-4: Mechanical Structure_ Side View ......................................................................... 49

Figure 7-5: Trays with Plantation ............................................................................................ 51

Figure 7-6: Sprinkler System ................................................................................................... 52

Figure 7-7: Constructed Controller .......................................................................................... 52

ix | P a g e

Figure 7-8: Constructed Mechanical Structure ........................................................................ 53

LIST OF TABLES

Table 3-1:Plant Type Categories ............................................................................................. 18

Table 3-2:Different Roof Types For simulation ..................................................................... 22

Table 3-3:Simulation Results ( Colombo) ............................................................................... 23

Table 3-4:Simulation Results (Anuradhapura) ........................................................................ 23

Table 3-5:Simulation Results (Hambantota) ........................................................................... 24

Table 3-6: Conclusion Table .................................................................................................... 27

Table 5-1:Arduino Board Specs. .............................................................................................. 40

Table 5-2: Solenoid Valve Specs ............................................................................................. 42

Table 5-3:Flow Sensor Specs................................................................................................... 43

Table 9-1: Part of Weather and Field Data Log ...................................................................... xxi

Table 9-2: Part of Irrigation Log ............................................................................................. xxi

1 | P a g e

1. INTRODUCTION

1.1. Green Roofs

The world is currently at a state of a crisis as its natural resources are being consumed at a rapid

rate by the ever increasing world population. The resultant pollutants of those consumption

activities such as greenhouse gases have become a threat to the survival of the planet. Therefore

there is an urgent necessity to adopt cleaner and greener technologies that are more

environmentally friendly as they reduce energy consumption and air and water pollution. A

green roof which is a vegetative layer grown on a roof providing shade and remove heat from

the air through evapotranspiration, reducing temperatures of the roof surface and the

surrounding air , is one of the cleaner and greener technology.

1.2. Green Roof Types

Green roofs can be classified in to two main categories, i.e. intensive green roofs and extensive

green roofs. Intensive green roofs have bigger substrate depths (20 – 60cm) and are able to

carry plants like shrubs and trees with deeper rooting systems. They are heavier (290 –

968kg/m2) and thus require enhanced structural capacity and higher capital cost and

maintenance. Extensive green roofs have thinner substrate layers (typically 5 – 15cm) and

lighter weight (approx. 72 – 169kg/m2). Their lower capital cost, low maintenance cost and

flexibility to install as a retrofit to older buildings with minimal structural changes make them

the green roof of choice. Extensive green roofs are mostly designed for natural precipitation

but may require additional irrigation during severe drought conditions. [1]

2 | P a g e

Different layers of the green roof are as follows.

Figure 1-1 Green Roof Layers

1.3. Green Roof Benefits

Green roofs provide multiple benefits to the buildings and environment.

2. Storm water Management

Green roofs mitigate the effects of storm water by retaining a portion of the rainfall they

capture. They also slow the flow of water that does come off the roof. These effects minimize

the amount of water coming from the rooftop into the storm sewers and ultimately into the

bodies of water that collect that water.

Many towns have a combined storm water or sanitary sewer system. In these cases all the

water that enters the sewer must be treated in a wastewater treatment plant, and when heavy

rains happen the treatment plant may overflow by releasing raw sewage into waterways. One

way to deal with this issue is to install a new sewer system or water treatment plants that are

sized to handle the maximum possible storm flow. Green roofs help mitigate this need, most

towns account for green roofs as permeable surface, the same as a grass field. So as towns

expand, green roofs help minimize the impact on storm water infrastructure [1].

3. Longer Life for the Roof Membrane

The water proofing material on a roof will degrade over time. A typical commercial roof may

need to be replaced as soon as 15 or 20 years. The degradation is caused by two factors, photo

degradation from sunlight and mechanical degradation from temperature extremes. A green

roof minimizes both these effects. Sunlight does not reach the membrane beneath a green roof.

And the temperature extremes that are typically experienced and greatly moderated by a green

roof.

3 | P a g e

4. Lower Energy Costs

Although research is preliminary, there is much evidence that green roofs help reduce the

flow of heat into and out of the surfaces they cover. There is also evidence that the benefit is

much greater in summertime because the reflectivity of the leaves of the plants and the

evaporation rates are maximized in the summer. The effect is also greater in lower, wider

buildings where the proportion of the building envelope covered by the green roof is greater,

than it is in taller, narrower buildings [1].

5. Mitigation of the Urban Heat Island Effect

Cities are hotter in the summer than nearby rural areas. The clustered buildings and paved

areas hold heat and release it slowly, resulting in an overall increase in temperature. Urban

Heat Island effects effect comfort of city residents as well as public health. Air conditioners

consume more electricity and actually increase the effect by pumping heat from indoors to

outdoors, inefficiently. City green spaces such as parks, tree-lined streets and green-planted

areas are cooler due to evapotranspiration (evaporation plus the transpiration of the plants

releasing moisture through photosynthesis). Research suggests that any increases in urban

green space, including green roofs, will help mitigate the Urban Heat Island Effect [2].

6. Habitats

Green Roofs are not similar to native landscapes. So they attract a number of insects and birds.

Native bees and honeybees love the pollen from Sedum and Delosperma. Many butterfly

species can be spied on green roofs. Ground nesting birds such as Killdeer have been known

to nest on green roofs.

7. Amenity and Aesthetics

Most well maintained extensive green roofs are pleasant places to be, especially in an urban

cityscape. [2] The cooler temperatures, lovely carpets of flowers, and butterflies they attract

make them valuable as amenity spaces for the people working or living within the building.

This can not only increase the productivity of employees, it can help in recruiting a higher

caliber of employee who would have an appreciation for all the benefits of a green roof.The

Need of irrigation for a green roof

Green roof environment is harsh since they expose to direct solar radiation, high wind speeds,

extreme heat and shallow substrate depths. Therefore it can hold limited amount of moisture

for plants survival.There is a need to give more emphasis on the green roof irrigation scheme

in order to optimize the benefits offered by the green roofs. Water is the key ingredient for

4 | P a g e

green roof optimization for plant survivability, plant cover, Enhancing biodiversity and

aesthetics, Soil evaporation and transpiration and Soil thermal conductivity. Therefore deeper

optimization on green roof irrigation is important.

The characteristics of an optimized irrigation system which is designed and implemented as

follows.

Provides adequate irrigation to maintain plants healthy

Does not provide excess water

Able to forecast irrigation requirement

Can adapt to varying climate conditions

Our Problem Statement is “Development of an adaptable and optimized irrigation control

mechanism for green roofs for water conservation and sustained green roof

performance”

5 | P a g e

2. LITERATURE REVIEW

The water requirements of green roofs depend mainly on evapotranspiration (ET), soil

chemistry, and the plant’s maximum allowable depletion (MAD) [3]. Direct measurements of

root zone soil moisture, water application along with predicted evapotranspiration values can

be used in a soil water balance model that can significantly optimize irrigation efficiency.

In soil there are main three soil moisture conditions namely

Field Capacity (FC), Permanent Wilting Point (PWP) and Maximum allowable Depletion .

FC can be defined as the threshold point at which the soil pore water will be influenced by

gravity. Above field capacity, the gravitational force will overcome the capillary forces

suspending the moisture in the pores of the soil allowing for down movement of water in the

soil column below. Permanent Wilting Point is the soil moisture level at which plants can no

longer absorb water from the soil. Plant transpiration and direct evaporation will decrease the

moisture level in soil to a point below PWP and, in some cases, down to near dryness .The

point below field capacity where plants become stressed is called the Maximum Allowable

Depletion. The Maximum Allowable Depletion value is expressed as a percent of the available

water capacity [3].

Plants can uptake water from soil if the soil moisture is above permanent wilting

point. As the soil moisture approaches permanent wilting point, the plant will become

increasingly stressed as the soil pore water becomes depleted. The point below field capacity

where plants become stressed is called the maximum allowable depletion (MAD).So in order

to optimize the green roof performance by optimizing irrigation, soil moisture level should be

maintained in between Field Capacity and Maximum Allowable Depletion.

Lower soil moisture target = FC-(FC-PWP) ×MAD [3]

When soil water balance is considered, by ET, Runoff and draining, water is removed from

Figure 2-1: Soil Moisture Conditions

6 | P a g e

soil. ET is the water that is transpired out of the soil by the plant plus the amount of water lost

to evaporation. ET represents the rate of water consumed by the plant and lost by direct

evaporation. The factors that affect the ET rate include wind, temperature, relative humidity,

and solar radiation. The units for ET are mm/hour.

Based on the Penman Monteith model for ET estimations, ET is not measured directly for an

individual crop, but rather it is determined from a standard reference grass and then adjusted

for different crops and plants with a crop coefficient. The evapotranspiration for a reference

grass is referred to as the potential evapotranspiration (ET0). Potential evapotranspiration

values will vary regionally and seasonally and with time.

2.1. Penman-Monteith Equation for Estimating ET

It is described the steps needed to estimate reference evapotranspiration (ETref) for a 0.12 m

tall reference surface (ETos) and for a 0.50 m tall reference surface (ETrs) using hourly weather

data.

i. Data Requirements

Site characteristics including the latitude (+ for north and – for south), longitude (+ for west

and – for east) and elevation (m) above sea level must be input. The required weather data

includes hourly solar radiation (MJ m-2h-1), mean air temperature (oC), mean wind speed (m s-

1) and mean dew point temperature (oC). The air and dew point temperatures should be

measured at between 1.5 and 2.0 m height and the wind speed should be measured at 2.0 m

height. For wind speeds measured at some height other than 2.0 m, the wind speed at 2 m height

(u2) can be estimated as:

𝑢2 = 𝑢𝑧 (4.87

ln(67.8𝑧𝑤 − 5.42))

whereuz = wind speed (m s-1) at height zw (m) above the ground.

STEP 1: Extraterrestrial radiation (Ra) is calculated for each hour using the following equations

from Duffie and Beckman (1980).

GSC = solar constant in MJ m-2 min-1

GSC = 0.082

σ = Steffan-Boltzman constant in MJ m-2 h-1 K-4

σ = 2.04 × 10-10

7 | P a g e

φ = Latitude in radians converted from latitude (L) in degrees

𝜙 =𝜋𝐿

180

J = day of the year (1-366)

dr = correction for eccentricity of Earth’s orbit around the sun

𝑑𝑟 = 1 + 0.033 cos (2𝜋𝐽

365)

δ = Declination of the sun above the celestial equator in radians

𝛿 = 0.409 sin (2𝜋

365𝐽 − 1.39)

Lm = station longitude in degrees

Lz = longitude of the local time meridian

Lz = 120o for Pacific Standard Time

Sc = solar time correction for wobble in Earth’s rotation

𝑏 =2𝜋(𝐽 − 81)

364

(3)

𝑆𝑐 = 0.1645 sin(2𝑏) − 0.1255 cos(𝑏) − 0.025 sin(𝑏)

t = local standard time (h)

ω = hour angle in radians

(4)

𝜔 =𝜋

12[(𝑡 − 0.5) +

𝐿𝑠 − 𝐿𝑚

15− 12 + 𝑆𝑐]

(5)

8 | P a g e

ω1 = hour angle ½ hour before ω in radians

𝜔1 = 𝜔 − (1

2) (

𝜋

12)

ω2 = hour angle ½ hour after ω in radians

𝜔2 = 𝜔 − (1

2) (

𝜋

12)

θ = solar altitude angle in radians

𝑠𝑖𝑛𝜃 = (𝜔2 − 𝜔1) sin(𝜙) sin(𝛿) + cos(𝜙) cos(𝛿) (sin(𝜔2) − sin(𝜔1))

Ra = extraterrestrial radiation (MJ m-2 h-1)

Ra = (60GSC )drsinθ (9)

β = solar altitude in degrees

β= sin−1[sinφsinδ+cosφcosδcosω] (10)

STEP 2: Calculate the hourly net radiation (Rn) expected over grass in MJ m-2 h-1 using

equations from Allen et al. (1994).

Rso = clear sky total global solar radiation at the Earth’s surface in MJ m-2 h-1

Rso= Ra(0.75+2.0×10−5El ) (11)

whereEl = elevation above mean sea level (m)

es = saturation vapor pressure (kPa) at the mean hourly air temperature (T) in oC

𝑒𝑠 = 0.6108𝑒[17.27𝑇

𝑇+237.3]

ea = actual vapor pressure or saturation vapor pressure (kPa) at the mean dew point

temperature

𝑒𝑎 = 0.6108𝑒[

17.27𝑇𝑑𝑇𝑑+237.3

]

ε′ = apparent ‘net’ clear sky emissivity

휀′ = 0.34 − 0.14√𝑒𝑎

Note thatε′=εvs−εa, where εvs is the emissivity of the grass and εa is the emissivity from the

atmosphere. It is called ‘apparent’ because the temperature from a standard shelter rather than

the surface temperature and atmosphere temperature are used to calculate the ‘net’ long–wave

9 | P a g e

radiation balance. Equation 11 is called the ‘Brunt form’ equation for net emittance because

the form of the equation is similar to Brunt’s equation for apparent long-wave emissivity from

a clear sky.

f = a cloudiness function of RS and RSO

𝑓 =1.35 𝑅𝑠

𝑅𝑠0− 0.35

with the restriction that 0.3 <Rs/Rso≤ 1.0 and Rs/Rso =0 whenever β<17.2o (=0.300 radians)

above the horizon. When using a spreadsheet program, put the value f = 0.6 in the cell before

the first data cell in the column for f. For each sequential hour interval, whenever β< 17.2o, let

the value for f equal the previous f value. When the corresponding β≥17.2o, use the Rs/Rso and

Equation 15 to calculate the f values. The values for f will fall between 0.05 and 1.00. If this

procedure is followed, the nighttime values for f will equal the f value at the end of the previous

daylight period until the next daylight period. The nighttimef values are used to estimate the

effect of cloud cover on Rn during the night. This method is used in the PMhr.xls program.

Rns = net short wave radiation as a function of measured solar radiation (Rs) in MJ m-2 h-1

Rns=(1−0.23)Rs (16)

To convert Rs from W m-2 to MJ m-2 h-1, multiply by 0.0036.

Rnl = net long wave radiation in MJ m-2 h-1

𝑅𝑛𝑙 = −𝑓휀′𝜎(𝑇 + 273.15)4

Rn = net radiation over grass in MJ m-2 h-1

Rn= Rns+ Rnl (18)

STEP 3: Calculate ETo using the Penman-Monteith equation as presented by Allen et al. (1994)

Bp = barometric pressure in kPa as a function of elevation (El) in meters

𝐵𝑃 = 101.3 (293 − 0.0065𝐸1

293)

λ = latent heat of vaporization in (MJ kg-1)

λ=2.45

10 | P a g e

γ = psychrometric constant in kPaoC-1

𝛾 = 0.00163 ×𝐵𝑝

𝜆

ra = aerodynamic resistance in s m-1 is estimated for a 0.12 m tall crop as a function of wind

speed (u2) in m s-1 as:

𝑟𝑎 =208

𝑢2

Modified psychrometric constant (γ∗)

For the short 0.12 m tall canopy during daylight (when Rn> 0), a canopy resistance of rs = 50 s

m-1 and an aerodynamic resistance of ra = 208/u2 are used to calculate modified psychrometric

constant as:

𝛾∗ = 𝛾 (1 +𝑟𝑠

𝑟𝑎) ≈ 𝛾(1 + 0.24𝑢2)

During the night (when Rn≤ 0), a canopy resistance of rs = 200 s m-1 and an aerodynamic

resistance of ra = 208/u2γ* are used to calculate the modified psychrometric constant as:

For wind speeds less than 0.5 m s-1, the wind speed is set equal to 0.5 m s-1 for both Eqs. 23

and 24. For the 0.50 m tall canopy during daylight (when Rn> 0), a canopy resistance of rs =

30 s m-1 and an aerodynamic resistance of ra = 118/u2 s m-1 are used to calculate the modified

psychrometric constant as:

𝛾∗ = 𝛾 (1 +𝑟𝑠

𝑟𝑎) ≈ 𝛾(1 + 0.25𝑢2)

During the night (when Rn≤ 0), a canopy resistance of rs = 200 s m-1 and an aerodynamic

resistance of ra = 118/u2 s m-1 are used to calculate the modified psychrometric constant as:

𝛾∗ = 𝛾 (1 +𝑟𝑠

𝑟𝑎) ≈ 𝛾(1 + 1.7𝑢2)

For wind speeds less than 0.5 m s-1, the wind speed is set equal to 0.5 m s-1 for both Eqs. 25

and 26.

∆ = slope of the saturation vapor pressure curve (kPaoC-1) at mean air temperature (T)

Δ =4099𝑒𝑠

(𝑇 + 237.3)2

11 | P a g e

G = soil heat flux density (MJ m-2 h-1)

For ETos, let G = 0.1 Rn when Rn> 0 and let G = 0.5 Rn for Rn< 0. For ETrs, let G = 0.04 Rn when Rn> 0 and G = 0.2 Rn when Rn ≤ 0.

R is the radiation term of the Penman-Monteith and Penman equations in mm d-1 .

When Rn> 0, for ETos, the radiation term contribution to ET is calculated as:

𝑅𝑜 =0.408Δ(𝑅𝑛 − 𝐺)

Δ + γ(1 + 0.24U2)

And during the night, it is calculated as:

𝑅𝑜 =0.408Δ(𝑅𝑛 − 𝐺)

Δ + γ(1 + 0.96U2)

When Rn> 0, for ETrs, the radiation term contribution to ET is calculated as:

𝑅𝑜 =0.408Δ(𝑅𝑛 − 𝐺)

Δ + γ(1 + 0.25U2)


𝑅𝑜 =0.408Δ(𝑅𝑛 − 𝐺)

Δ + γ(1 + 1.7U2)

For the ETp(Penman equation), the radiation term contribution to ET is calculated as:

𝑅𝑜 =0.408Δ(𝑅𝑛 − 𝐺)

Δ + γ

for both day and night calculations.

A = aerodynamic term of the Penman-Monteith equation in mm d-1 with u2 the wind speed at

2 m height

When Rn> 0, for ETos, the aerodynamic contribution to ET is calculated as:

𝐴𝑜 =(

37𝛾

𝑇𝑀+273) 𝑢2(𝑒𝑠 − 𝑒𝑎)

Δ + γ(1 + 0.24u2)

12 | P a g e


𝐴𝑜 =(

37𝛾

𝑇𝑀+273) 𝑢2(𝑒𝑠 − 𝑒𝑎)

Δ + γ(1 + 0.96u2)

When Rn> 0, for ETrs, the aerodynamic contribution to ET is calculated as:

𝐴𝑟 =(

66𝛾

𝑇𝑀+273) 𝑢2(𝑒𝑠 − 𝑒𝑎)

Δ + γ(1 + 0.25u2)


𝐴𝑟 =(

66𝛾

𝑇𝑀+273) 𝑢2(𝑒𝑠 − 𝑒𝑎)

Δ + γ(1 + 1.7u2)

For ETp, the aerodynamic contribution to ET during daytime and night time is calculated as:

𝐴𝑝 =(

37𝛾

𝑇𝑀+273) 𝑢2(𝑒𝑠 − 𝑒𝑎)

Δ + γ

Reference evapotranspiration

For a short (0.12 m) canopy, the Penman-Monteith reference evapotranspiration is calculated as:

ETos= Ro + Ao (38)

Similarly, for a tall (0.5 m) canopy, the Penman-Montieth reference evapotranspiration is calculated as:

ETrs= Rr+ Ar (39)

For a short (0.12 m) tall canopy, the Penman evapotranspiration is calculated as:

ETos= Ro + Ao (40)

In equations 38-40, the units are mm h-1 [4]

2.2. Irrigation scheduling methods

Initially the existing irrigation scheduling methods are studied. Because of the variability and

13 | P a g e

irregularity in rainfall make irrigation scheduling difficult. Therefore an efficient irrigation

schedule which is applying the right amount of water at the right time is essential in meeting

the dual goals of water conservation and sustained green roof performance. Under-irrigation

and over-irrigation can negatively affect benefits of green roof and over-irrigation results in

waste of water and leaching of nutrients.

So-called “smart” irrigation technologies can be separated into two categories as those that use

a feedback sensor to monitor the amount of water in the root zone, and those that use weather

data and a soil-water budget to adjust irrigation amounts.

Systems that are based upon soil-water feedback incorporate buried sensors that are wired in

series with and electrically actuated solenoid valve. The sensor acts as a switch opening the

circuit between the controller and the valve when the soil-water content is high preventing

irrigation, and closing the circuit when watering is needed.

Another type of system used in irrigation control is based on controllers that use weather

information to estimate ET and adjust irrigation using a soil-water budget. An ET controller

can make adjustments to the watering schedule based on Weather conditions without requiring

human interaction. ET controllers receive information from local or on-site weather stations

and adjust watering durations to replace ET.

Standard timer-based irrigation

This type of system set to apply water on a fixed schedule (weekly, bi-weekly, and daily) and

irrigation duration to replace the historical irrigation requirement (adjusted monthly). The

irrigation requirement factored in effective precipitation and the irrigation system uniformity.

Soil moisture-based “add-on” system

Soil-moisture feedback sensors were placed in block 2 plots of each irrigation frequency for

the add-on system. These modules were connected to a standard irrigation controller with three

independent programs similar to the time based controller. The controller was set to apply the

same amount of water as the standard time-based system.The “add-on”system has a Time

Domain Transmissivity (TDT) moisture sensor that measures the volumetric moisture

percentage of the soil and prevents irrigation above a user-supplied soil-water content. The

volumetric soil-water setpoint used in this study was 24%, equivalent to 75% of field capacity

per manufacturer directions.

Soil moisture based on-demand system

The “on-demand” soil-water feedback controller system evaluated uses the same sensor as the

14 | P a g e

“add-on” system; however it is designed to both initiate and terminate irrigation events by

setting lower and upper setpoints respectively. The lower and upper setpoints were set at 21%

and 30% volumetric moisture, respectively, with the lower Setpoint (turn-on) corresponding to

a depletion of 67% of the plant-available soil water, and the upper setpoint (turn-off) just below

field capacity. Cycle and soak times of 10 minutes were programmed to allow for water

infiltration and sensor detection.

Evapotranspiration-based system

The ET-based controller was evaluated at the same irrigation frequencies as the timer-based

and “add-on” systems. The plots irrigated by the ET controller system received irrigation

amounts based upon reference ET estimates downloaded daily from a weather service provider

and a soil-water budget. User inputs that affect the soil-water budget include root depth, soil

type, crop type, and sun exposure. [5]

2.3. Sprinkler placement for uniformly distribution of water

The area watered by the adjacent sprinkler should be overlapped substantially by the area

watered by each sprinkler. This overlap may get seemed like a waste at first, but it is a quite

important necessity. It would not be possible to design sprinkler system that provided uniform

water coverage without this overlap. If there is no 100% overlap in the watering area then there

will be dry points. This method is called head to head coverage or head to head spacing. [6]

Decide which edges to start placing sprinklers

First, decide which sides of the yard must have sprinklers along the edge. For example, hard

edges like curbs, driveways and also foundations usually have sprinklers adjacent to the edge.

For green roof orientation it is more effective to place sprinklers along every edge to elude

watering beyond the roof edge line, and to provide optimum overlap at all points along the

edge, but this also increases the installation cost somewhat higher. Placing a row of full circle

sprinklers about one quarter to one sixth of the diameter from an edge will not water quite as

evenly along the edge and will throw water beyond the green roof edge or if it is undergrowth

area on the land then watering beyond the fence or property , but it decreases the installation

cost slightly. If you have bushes along the edge of green roof or vegetation on land then along

the fence and desire to water them at the same time, then the sprinklers must be placed several

feet in front of the shrubs rather than behind them. If the sprinklers are placed behind the bushes

and must throw water over them, then the spray will be obstructed when the bushes grow larger,

15 | P a g e

ensuing in terrible watering uniformity. [6]

There are two types of sprinkler spacing which are called square spacing and triangular spacing.

Following figure shows both triangular spacing and square spacing. [6]

Figure 2-2:Types of sprinkler Spacing

Figure 2-3:triangular Spacing

Figure 2-4:Square Spacing

Determination of this spacing types are as follows.

a. First, measure the length and width of the largest open areas (from

the sprinkler locations you chose near the edges above) to see how many rows of

sprinklers are required.

b. Look at the shape of the open area to see if there is an obvious choice between square

or triangular spacing. Use square spacing unless there is a reason to use triangular

spacing. If roof top corners or the land corners are near 60° or 120° you may find

triangular spacing which fits better.

c. One more reason to use triangular spacing may be if an excessive amount of row

compression is required to make square spacing fit. For example, if the rows must be

compressed more than 0.86 times the head spacing, you should think about triangular

spacing, as this automatically favours layout rows closer than heads within a row.

16 | P a g e

Determining rotor nozzle sizes

Since some of rotors may be watering a full circle on the green roof top or land, meanwhile

other rotors will be watering a part of a circle, it is vital to factor this into the flow rate for

the nozzle selected for each sprinkler. When the area increases, the radius (throw distance) for

a nozzle is relatively independent of the nozzle flow rate. A sprinkler watering a full circle

should extinguish four times watering as much as a sprinkler in the corner of the roof top or

the yard watering only a quarter circle. Similarly, a sprinkler in the mid of a long straight edge

should put out half as much water as the full circle sprinkler. If this is not considered to

influence into your nozzle selection, there won't have an uneven watering, and unnecessary

money will be consumed on watering that is wasted. [6]

http://www.watertips.com/info/glossary.htm#nozzle

17 | P a g e

3. STUDY ON THE NEED FOR OPTIMIZED IRRIGATION FOR

GREEN ROOFS IN DIFFERENT ZONES IN SRI LANKA

3.1. Overview

Objective of the study is to develop a water balance model to estimate irrigation requirements

for green roofs in different zones in Sri Lanka

3.2. Simulation

Simulation has been done to develop an understanding about the average water availability/

deficit in different zones in Sri Lanka. This results help us to develop a water balance model

that can act as our base for future studies. One year weather data from 3 weather stations (Wet

zone, dry zone and intermediate zone) is used for the study. Data used is as follows

o Daily rainfall

o Daily max. and min. temperatures

o Daily average wind speed

o Daily average relative humidity

o Daily average solar radiation

One of our project objectives is to implement the irrigation control system for green roofs in

different climatic zones in Sri Lanka. Before implementing, justification for the requirement

of artificial irrigation is needed. We simulate different green roof types and there water

balance without artificial irrigation. To obtain that, following methodology is used.

Figure 3-1 Green Roof simulation Method

18 | P a g e

1. Obtained one year weather data as mentioned below from 4 weather stations in Sri

Lanka. Those are Colombo, Anuradhapura, Kurunegala and Hambantota representing

Wet zone, dry zone, intermediate zone and semi -arid zone respectively.

o Daily rainfall

o Daily max. and min. temperatures

o Daily average wind speed

o Daily average relative humidity

o Daily average solar radiation

2. There are several plant categories available in Sri Lanka mainly CAM (Crassulacean

Acid Metabolism) and C4 type. Plant characteristics of these two types are mentioned

in the following table.

Table 3-1:Plant Type Categories

CAM type C4 type

Low mat forming or compact growth Photosynthesis saturates at lower CO2

concentrations than C3 plant

Evergreen foliage Not as draught tolerant as CAM plants

Fewer stomata

CO2 synthesis during night

Tough, twiggy growth

Succulent leaves and water storage

capacity

Examples are Cactus and Pineapple Examples are UmariKeerai,

Citronella, Lemongrass.

For the simulation purpose, Succulant Grass as CAM type and Grass Herbs as C4 type

are used.

3. To calculate Evapo-transpiration (i.e. combination of Evaporation and Transpiration),

ETo calculator is a software developed by the Land and Water Division of FAO.

Main function of the calculator is to calculate Reference evapo-transpiration

(ETo)according to FAO standards. Inputs to the ETO are annual data files of daily

19 | P a g e

average wind speed, average solar radiation, daily average temperature, daily rainfall

and daily average relative humidity. The theory behind the calculation is the Penman-

Montieth equation. Penman–Monteith equation is sensitive to vegetation specific

parameters, but in this calculation they were considered as insignificant. Equation is

given below mentioning all the parameters. [7]

𝐸𝑇𝑜 =∆𝑅𝑛 + 𝜌𝑎𝐶𝑝(𝛿𝑒)𝑔𝑎

[∆ + 𝛾(1 +𝑔𝑎

𝑔𝑠)]𝐿𝑣

Then the ETO file is obtained for simulation in the green roof water balance. Following

graphs shows the daily evapo-transpiration in each zone throughout the year.

Wet zone (Colombo)

Figure 3-2:Daily ET of Colombo (2012)

𝐿𝑉 − Volumetric latent heat of vaporization (2453MJ/m3)

𝐸𝑇𝑜 – Water volume evapotranspired (mm/s)

∆ – Rate of change saturation specific humidity and temperature (Pa/K)

𝑅𝑛– Net irradiance (W/m2)

𝐶𝑝 – Specific heat capacity of air (J kg-1K-1)

𝛿𝑒- Vapor pressure deficit or specific humidity (Pa)

𝑔𝑠 −Conductivity of Stoma, surface conductance (ms-1)

𝜌𝑎 − Dry air density (kg m-3)

𝑔𝑎 − Conductivity of air (ms-1)

20 | P a g e

Dry zone (Anuradhapura)

Semi- Arid zone ( Hambantota)

Green roof water balance is considered as follows.

Incoming water streams = Rainfall + Artificial irrigation.

Outgoing water streams = Evapo-transpiration + Surface runoff

Incoming water streams = Outgoing water streams + Water storage in the soil

The Green Roof model considers the water balance of roofs. The amount of water

(i) Retained on the roofs

(ii) Evacuated as run-off from the roofs

Figure 3-3:Daily ET of Anutadhapura (2012)

Figure 3-4:Daily ET of Hambantota (2012)

21 | P a g e

are computed on a daily basis for a simulation period. When the weather conditions are

favourable, rain water retained on the roof is removed by Evapotranspiration. The input

contains files containing daily climatic data over a year for the location and characteristics for

the studied roof. Gravel roofs, tile and slate roofs, bitumen roofs and extensive green-roofs

with numerous types of vegetation and degree of vegetation cover can be selected or can be

formed and saved in the data bank for later use. The roofs are characterized according to their

surface area, position, orientation, and slope and for green-roofs by the type and extension of

the vegetation cover, depth of the substrate layer and the presence of a drainage layer. The

climatic data contains daily rainfall observed in a representative weather station and grass

reference Evapotranspiration (ETo). The grass reference evapotranspiration characterises the

evaporative demand of the atmosphere for the location and is derived from meteorological data.

The water balance of the green roof is computed with a daily time step by keeping track of the

daily incoming and outgoing water fluxes. Since roofs oriented to a dominant wind direction

will receive more rainfall than flat roofs or oriented to another direction, the model estimate

the incoming water by considering not only rainfall but also the slope and the orientation of the

roof. [8]

3.3. Amount of water loss due to ET

This ET denotes the evaporation of water on the roof or vegetation surfaces and the

transpiration of water by vegetation stored in the substrate of green-roofs. The quantity of water

vanished due to the ET is determined by following parameters.

I. Reference ETO

II. The slope and the orientation of the green roof.

III. Characteristics of the roof surface

IV. Surface area

V. Amount of the water retained by the roof.

In this software ET is given as input with unit of mm/day. Even though ETo is not very sensitive

to wind speed, a correction is required if the wind speed observed on the roof powerfully

deviates from the observations in the meteorological station.

For identical environmental conditions, the ET from a wet green roof varies from ETo due to

differences in the characteristics of the roof surface when compared with the reference grass

surface. By incorporating the effect of evaporation and transpiration into a single ‘crop’

coefficient (Kc), the ET from a wet roof is given by the Kc approach (Allen et al., 1998):

22 | P a g e

ET =Kc *ETO

Where ETO (mm.day-1) is the reference evapotranspiration modified and adjusted for slope,

orientation and position if required. By choosing a roof type, the program considers the

appropriate Kc coefficient. The Kc of green-roofs depends on the type of vegetation and degree

of soil cover. The Kc coefficient can be adjusted by the user to consider local conditions.

The inputs given to this green roof simulator software are annual daily ET file taken from ETO

calculator, plant type, Surface area, inclination type and Substrate depth of the roof. Following

parameters are changed in the software for simulation. [8]

• Succulent – Grasses: CAM type plants

70% ET with respect to reference ET

50% max. ET without water stress

• Grasses – Herbs: Typically C4 type

100% ET with respect to reference ET

50% max. ET without water stress

Simulation was implemented for following green roof designs. There were 12 green roof

designs.

Table 3-2:Different Roof Types For simulation

Vegetation Vegetation

cover (%)

Surface

Area(m2)

Inclin

ation

Substrate

depth(cm)

Orienta

tion

Exposure

Succulents- grass 25 50 Flat 5 NA Fully exposed

Succulents- grass 50 50 Flat 5 NA Fully exposed

Grass-Herbs 25 50 Flat 5 NA Fully exposed

Grass-Herbs 50 50 Flat 5 NA Fully exposed

Succulents- grass 25 50 30o 5 East Fully exposed

Succulents- grass 50 50 30o 5 East Fully exposed

Succulents- grass 25 50 30o 5 West Fully exposed

Succulents- grass 50 50 30o 5 West Fully exposed

Grass-Herbs 25 50 30o 5 East Fully exposed

Grass-Herbs 50 50 30o 5 East Fully exposed

Grass-Herbs 25 50 30o 5 West Fully exposed

Grass-Herbs 50 50 30o 5 West Fully exposed

This simulation was implemented for four different climatic zones as mentioned above in Sri

23 | P a g e

Lanka. Results of water balance for these climatic zones are summarized as follows. This

shows the annual water balance without artificial irrigation.

Colombo (Wet zone)

Table 3-3:Simulation Results ( Colombo)

Vegetation Vegeta

tion

cover

(%)

Inclinati

on

Total

rainfall(liter)

Total evapo-

transpiration

Water storage

(liter)

Runoff (liter) Water

stress days

Stress %

Succulents- grass 25 Flat 118960 52644.6 362.7 65952.7 252 68.85


Grass-Herbs 25 Flat 118960 54342.6 288.6 64328.8 257 70.22

Grass-Herbs 50 Flat 118960 53318.1 317.3 65324.5 255 69.67

Succulents- grass 25 30o 89423.4 43579 72.3 45772.1 267 72.95

Succulents- grass 50 30o 89423.4 41326.7 150.9 47945.7 260 71.04



Grass-Herbs 25 30o 89423.4 44971.6 47.3 44404.5 276 75.41

Grass-Herbs 50 30o 89423.4 44168.4 86.6 45168.5 274 74.86

Grass-Herbs 25 30o 116621.3 50721 21.5 65878.8 284 77.60

Grass-Herbs 50 30o 116621.3 49795.1 52.2 66774 281 76.78

Anuradhapura (Dry Zone)

Table 3-4:Simulation Results (Anuradhapura)

Vegetation Vegeta

tion

cover

(%)

Inclinati

on

Total

rainfall(liter)

Total evapo-

transpiration

Water storage

(liter)


stress days

Stress %

Succulents- grass 25 Flat 96480 28817.8 314.8 67977 314 85.8


Grass-Herbs 25 Flat 96480 29700.3 359.6 67139.3 315 86.1

Grass-Herbs 50 Flat 96480 29291.3 316.7 67505.4 314 85.8

Succulents- grass 25 30o 29010 9442 121.8 19689.8 321 87.7

Succulents- grass 50 30o 72525 25293.3 248.4 47480.4 310 84.7

24 | P a g e


Succulents- grass 50 30o 94583.3 28341 280.4 66522.6 312 85.2

Grass-Herbs 25 30o 72525 27211.9 386.7 45699.8 320 87.4

Grass-Herbs 50 30o 72525 26847.3 344.5 45022.2 317 86.6

Grass-Herbs 25 30o 94583.3 30558.5 400.7 64425.5 318 86.9

Grass-Herbs 50 30o 94583.3 30190.3 361.1 64754.1 316 86.3

Hambantota (Semi Arid Zone)

Table 3-5:Simulation Results (Hambantota)

Vegetation Vegeta

tion

cover

(%)

Inclinati

on

Total

rainfall(liter)

Total evapo-

transpiration

Water storage

(liter)


stress days

Stress %



Grass-Herbs 25 Flat 62590 36129.1 420.2 26881.1 332 90.7

Grass-Herbs 50 Flat 62590 35669.8 394.4 27314.6 330 90.2


Succulents- grass 50 30o 47049.5 30296.8 392.7 17145.4 333 91



Grass-Herbs 25 30o 47049.5 31568 426.2 15907.7 337 92.1

Grass-Herbs 50 30o 47049.5 31214.6 404.9 16239.8 337 92.1

Grass-Herbs 25 30o 61359.5 36482 425.4 25302.9 333 91

Grass-Herbs 50 30o 61359.5 36036 403.3 25726.8 333 91

According to these simulation results, highest water stress roof type was selected and then we

obtained daily water balance of these four climatic zones.

25 | P a g e

Wet zone (Colombo)

Figure 3-5:Daily Water Balance (Colombo)

Intermediate zone (Kurunegala)

Figure 3-6:Daily Wataer Balance (Kurunegala)

0

100

200

300

400

500

600

700

800

900

-34 16 66 116 166 216 266 316 366

Soil Moisture Level (Liter)s

Days

COLOMBO_2012

Water Retention

Field Capacity

Lower SoilMoisture Target

0

100

200

300

400

500

600

700

800

900

-34 16 66 116 166 216 266 316 366

Soil Moisture Level (Liters)

Days

KURUNEGALA_2012

Water Retention

Field Capacity

Lower SoilMoisture Target

26 | P a g e

Dry zone (Anuradhapura)

Figure 3-7:Daily Water Balance (Kurunegala)

Semi- Arid zone ( Hambantota)

Figure 3-8:Daily Water Balance (Hambantota)

According to these water balance results we came up with following conclusions.

Water retention of the substrate is below the lowest soil moisture target is as follows in each

zones.

0

200

400

600

800

1000

0 100 200 300


Days

ANURADHAPURA_2012

Water Retention

Field Capacity

Lowest SoilMoisture Target

-200

0

200

400

600

800

1000

-34 66 166 266 366


Days

HAMBANTOTA_2012

WaterRetention

Field Capacity

Lower SoilMoistureTarget

27 | P a g e

Table 3-6: Conclusion Table

Zone Water retention

Wet zone (Colombo) About 70% of the year

Intermediate zone ( Kurunegala) About 85% of the year

Dry zone ( Anuradhapura) Almost entire year

Semi- Arid zone ( Hambantota) Almost entire year

3.4. Conclusion of the Study for the justification of the artificial irrigation

requirement.

Both plants stress more than 70% throughout the year in all zones.

Our emphasis is to improve the plant performance without stress and save the use of

excess water.

So, artificial irrigation is implemented as required for all zones.

28 | P a g e

4. DEVELOP A CONTROL ALGORITHM TO IMPLEMENT THE

IRRIGATION SCHEME

4.1. Overview

When soil water balance is considered , precipitation and irrigation adds water to soil and

Evapotraspiration ,Draining and run off removes the water from soil. In an efficient irrigation control

system keeps draining and run off insignificant .Therefore only factor that affects the removal of

water from the soil is Evapotransipration. The following diagram shows the soil water balance.

Irrigation control Algaorithm is developed using the soil water balance.

Figure 4-1:Evapotranspiration Overview

In the control algorithm, main three factors,current soil moisture content , Diffusion time and

Evapotranspiration. are considered. In this system sole soil moisture based closed loop system

is not used due to the consideration of removing the water from evapotraspiration during the

diffusion time. If sole soil moisture based closed loop system is used , compensation during

the diffusion time is not taken in to consideration. There is significant amount of water is

removed from soil by evapotraspitaion during the diffusion time. Therefore it not possible to

achieve the required soil moisture level by using sole soil moisture based irrigation control. So

in this control system compensation water volume during diffusion time is also added to the

(FC-MAD) water volume.

29 | P a g e

Irrigation control by predicting the evapotranspitaion for the current hour.

Hourly evapotranspiration value is calculated by using the penman Montieth Equation. For the

calculation of the hourly evapotranspiration, hourly average data of temperature

,humidity,solar radiation and wind values are required. In order to get the current value of

evapotranspiration , prediction method is used(Obtaining current ET value by past hourly ET

values.).

Control algorithm is described step by step in the following.

STEP 01. Measuring Temperature , Humidity ,Wind and Solar Radiation

Instantaneous Temperature and Humidity is measured using DHT11 sensor and instantaneous

wind speed is measured using the anemometer. Solar radiation calculation is used to get hourly

solar radiation values.

STEP 02. Calculating hourly average values of temperature , humidity and wind using the

sensor readings

STEP 03. Predicting the Evapotranspiration value for the current hour

Modified seasonal decomposition method is used to predict the evapotranspiration value.

STEP 04. If current soil moisture level is less than the MAD , irrigating a volume of (Field

Capacity-Lowest soil moisture target) + Compensation for diffusion time

STEP 05. Waiting for the diffusion.

Diffusion time is estimated experimentally using the test bed for given substrate.

Algorithm is further explained using the following flow chart

30 | P a g e

Figure 4-2: Overall algorithm

31 | P a g e

Predicting Evapotranspiration

Before using the modified decomposition method which is used in the final Algorithm, several

prediction methods are used to predict ETo and checked the accuracy. Out of all the prediction

methods tested , modified seasonal decomposition method was the most accurate prediction

method that gives most approximate value for the current ETo value.

Average hourly Temperature and Humidity value for the current hour is predicted using the

seasonally decomposition method by past instantaneous data. Wind hourly average value for

current hour is predicted using moving average method. Current Solar radiation value is

calculated using the solar radiation calculation. Using these hourly predicted values

evapotranspiration is calculated for current hour.

Figure 4-3: Modified Seasonal Decomposition method

32 | P a g e

Figure 4-4: Seasonal Decomposition for Temperature

Figure 4-5:Seasonal Decomposition for Humidity

33 | P a g e

Figure 4-6:Seasonal Decomposition for solar radiation

Figure 4-7:Moving Average for Wind Speed

34 | P a g e

4.2. Programming

For the ease of coding the ETocalculation , separate library has been created called “ET”. The

following block diagram shows the libraries used in programming.

Figure 4-8:Libraries Used

Time.h : Synchronise the time and run a real time clock in Arduino.

TimeAlarms.h : Time tasks are created using this library.

DHT.h : DHT11 interfacing library.

ET :ET calculation library.

LiquidCrystal.h : Driver circuit for the LCD display.

Average.h : Library used for statistical data

Figure 4-9:Connectivity of Function

35 | P a g e

Above block diagram shows the connectivity between each functions in the code.

Lcd_update : Display shows the values of different parameters one after another

Serial Event : When input is reached to the Arduino , it is called.

Loop : Main loop

Setup : Setting input and output pins and setting up objects created by libraries

Lcd_out : Giving commands to display

Terminate_irr : Stop irrigation suddenly

Min_update : Update minute by minute

Checksync : Time synchronisation

Initialize : Reads the sensors when Aurdino is on

Wind_update : Reads the wind sensor

Soil_update : Reads the soil moisture sensor

Irr : Control of irrigation

Hour_update : Hourly average values are calculated

Serialsend : Transmit data to the PC

Predict_ET : Predict evapo-transpiration for current hour

readArialSensors : Read temperature and humidity of the atmosphere

wind_speed : Read the analogue value of anemometer

soil_moiture : Read the soil moisture sensor

s_out : Transmit serial data

Cal_J : Calculate day of the year

LPF : Software based low pass filter

36 | P a g e

5. ELECTRICAL DESIGN

5.1. Overall system

Figure 5-1: Overall System

5.2. Sensors that are used in local controller

01. DHT11 (Temperature and Humidity Sensor)

Measurement range for Humidity is 20-90 % and for temperature is 0-50 °C. Temperature

accuracy is ± 1 °C and humidity accuracy is ±4 %. A Stevenson shield which is an enclosure

to shield meteorological instruments against precipitation and direct heat radiation from outside

sources, while still allowing air to circulate freely around them is used to cover the DHT11

sensor. [9]

Figure 5-3:Stephenson Sheild Figure 5-2:DHT11 sensor

37 | P a g e

02. Anemometer

Anemometer is designed using a DC motor and amplifier circuit is used to amplify the voltage

given by the DC motor. Amplifier circuit was designed using NI Multisim Simulation software

Figure 5-4:Amplifier For Anemometer

Anemometer was calibrated to measure the instantaneous wind speed. Wind sensor was

calibrated by using following steps.

Anemometer was mounted on a car.

Driving the car at constant speeds on a flat road.

Taking down notes of the car's speed and the corresponding voltage read out by a multi

- meter

Graph the results and derive the proportional relation.

Calibration results are as follows.

wind

Speed(km/h)

Wind Speed

(m/s)

Voltage

(mV)

Amplifier Output

Voltage (V)

Analog

Read

Rounded

Value

20 5.555555556 50 0.5485 112.2231 112

30 8.333333333 120 1.3164 269.33544 269

40 11.11111111 150 1.6455 336.6693 337

50 13.88888889 215 2.35855 482.55933 483

60 16.66666667 250 2.7425 561.1155 561

38 | P a g e

Figure 5-5: Anemometer Calibration Results

03. Soil moisture sensor

Used soil moisture sensor is resistive type soil moisture sensor. It measures the soil

conductivity of the soil and it is proportional to the soil moisture content.

Figure 5-6: resistive type soil moisture sensor

It gives the analogue output. This contains a resistive type voltage divider which one is 10kΩ

and other one is the soil resistance measured.

y = 0.0304x

0

2

4

6

8

10

12

14

16

18

0 100 200 300 400 500 600

Win

d S

peed(m

/s)

Analog Read Value

Wind Sensor Calibration Results

39 | P a g e

We calibrated this soil moisture sensor in the following order.

i. Known amount of weight soil sample from the soil used for test bed was collected and

saturated that sample from the water.

ii. Sample was kept for 48 hours and sensor reading was read

iii. Weight of this wet sample was measured

iv. This wet sample was kept inside the oven at 100oC for 24 hours

v. Dry weight was measured

vi. Sensor reading of dry sample was read.

vii. Volumetric water content at field capacity was calculated

viii. By assuming sensor reading is linear, equation was obtained for the sensor

Calculation procedure

Sample volume – 50cm3

Table 5-1: calibrated results

Wet weight (g) Sensor reading Dry weight (g) Sensor

reading

Sample 1 60.1 852 40.60 25

Sample 2 60.3 854 40.50 22

Sample 3 60.2 853 40.45 22

Average values 60.2 853 40.55 23

𝑤𝑒𝑖𝑔ℎ𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑤𝑎𝑡𝑒𝑟 𝑖𝑛 𝑡ℎ𝑒 𝑠𝑎𝑚𝑝𝑙𝑒 = 65.2 − 40.55 = 24.65𝑔

𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑤𝑎𝑡𝑒𝑟 𝑖𝑛 𝑡ℎ𝑒 𝑠𝑎𝑚𝑝𝑙𝑒 = 24.65𝑐𝑚3

𝑣𝑜𝑙𝑢𝑚𝑒𝑡𝑟𝑖𝑐 𝑤𝑎𝑡𝑒𝑟 𝑐𝑜𝑛𝑡𝑒𝑛𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑠𝑜𝑖𝑙 =24.65

50= 0.493

𝑣𝑜𝑙𝑢𝑚𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑤𝑎𝑡𝑒𝑟 𝑜𝑓 𝑑𝑟𝑦 𝑤𝑒𝑖𝑔ℎ𝑡 = 0𝑔

𝑣𝑜𝑙𝑢𝑚𝑎𝑡𝑟𝑖𝑐 𝑤𝑎𝑡𝑒𝑟 𝑐𝑜𝑛𝑡𝑒𝑛𝑡 𝑎𝑡 𝑑𝑟𝑦 𝑠𝑜𝑖𝑙 = 0

𝑠𝑙𝑜𝑝𝑒 𝑜𝑓 𝑡ℎ𝑒 𝑙𝑖𝑛𝑒𝑎𝑟 𝑒𝑞𝑢𝑎𝑡𝑖𝑜𝑛 =0.493 − 0

853 − 23= 0.000594

𝑣𝑜𝑙𝑢𝑚𝑒𝑡𝑟𝑖𝑐 𝑤𝑎𝑡𝑒𝑟 𝑐𝑜𝑛𝑡𝑒𝑛𝑡 = 0.000594 ∗ 𝑠𝑒𝑛𝑠𝑜𝑟 𝑟𝑒𝑎𝑑𝑖𝑛𝑔

40 | P a g e

5.3. Irrigation controlling circuit

The ArduinoUno microcontroller board is used as the central controller. The ArduinoUno

micro-controller board is based on the ATmega328. It has 14 digital Input/output pins (of

which 6 can be used as PWM outputs), 6 analog inputs, a 16 MHz ceramic resonator, a USB

connection, a power jack, an ICSP header, and a reset button. [10]

Figure 5-7:Arduino Board

Arduino Board Controller Specification is as follows.

Table 5-2:Arduino Board Specs.

Microcontroller

Operating Voltage

Input Voltage (recommended)

Input Voltage (limits)

Digital I/O Pins

Analog Input Pins

DC Current per I/O Pin

DC Current for 3.3V Pin

Flash Memory

SRAM

EEPROM

Clock Speed

ATmega328

5V

7-12V

6-20V

14 (of which 6 provide PWM output)

6

40 mA

50 mA

32 KB (ATmega328) of which 0.5 KB used

by bootloader

2 KB (ATmega328)

1 KB (ATmega328)

16 MHz

41 | P a g e

Power supply Unit

Figure 5-8:Power Supply Unit Circuit Diagram

Figure 5-9: Power Supply Unit Hardware

42 | P a g e

5.4. Irrigation actuation

The operation of the actuator system is to control the amount of water quantity to the field. So,

there are two units that are used for actuation of the irrigation control system of the prototype.

They are solenoid valve and flow sensor. [11]

1. Solenoid valve

Figure 5-10: Solenoid Valve

This is connected to the water supply through flow sensor. Basic specifications are as follows.

Table 5-3: Solenoid Valve Specs

Model No AQT12SLT

Thread size 3/4"BSP inlet and 12mm outlet

Material Plastic and Brass

Voltage 220V AC

Style Open valve

Working environment Water, gas and oil

Solenoid valve is controlled through a relay. Outlet of the solenoid valve is connected to the

irrigation sprinkler system which is at the test bed. Basic function of the solenoid valve is to

control the water supply to the field as required through controller circuit. Controller circuit

of the solenoid valve is shown below.

Figure 5-11:Driver Circuit for Solenoid valve

43 | P a g e

Operation of the control circuit

CN10 receives the valve opening signal from the Aurdino then R4 transistor biased and VCE

becomes almost zero. Voltage across the coil is 9V DC and then it closes contactor of the circuit

of the solenoid valve. When the signal is removed from the CN10, voltage across the coil of

the relay is zero and hence contactor opened.

2. Flow sensor.

Water flow sensor consists of a plastic valve body, a water rotor, and a hall-effect sensor. When

water flows through the rotor, rotor rolls. Its speed changes with different rate of flow. The

hall-effect sensor outputs the corresponding pulse Signal. Three wires connecting to the sensor

are red, yellow and black which represent the 5V-24V DC supply, output pulse signal path and

ground terminal.

Figure 5-12:flow Sensor

Specification of the flow sensor is as follows. [12]

Table 5-4:Flow Sensor Specs.

Working voltage 5V-24V DC

Maximum current 15mA (5V DC)

Flow rate range 1-30 liters/min

Operating humidity 35%-90%

Operating temperature 0 oC to 80 oC

Operating pressure Under 1.2 MPa

Calibration result of the flow sensor is given in the following graph. At zero pulses volume of

water through the flow sensor is 37.02ml. So, to get accurate operation, irrigation of water

should be more than this value. Duty of the generated pulses is 0.5.

44 | P a g e

Figure 5-13:Flow Sensor Caliberation Results

5.5. Wireless communication between the local controller and PC

Hardware

Xbee S2 module is used for this purpose. This module is connected to the Aurdino via Xbee

shield and is connected to the PC via Xbee explorer. [13]

Figure 5-14:Xbee

Figure 5-15: Xbee Shield

Key features of Xbee S2 module

High performance

o Urban range/indoor- 40m and outdoor line of sight- 120m

o RF data rate- 250000 bps

o Serial interface data rate- 1200-230400 bps

-100

0

100

200

300

400

500

600

0 200 400 600 800 1000 1200

No of pulses

volume of water (ml)

CALIBRATION CURVE OF FLOW SENSOR

45 | P a g e

Low power consumption

o Supply voltage- 2.8V-3.4V

o Operating current 40mA for both transmission and receiving at 3.3V

Advanced networking and security

o Supported network topologies- point to point, point to multipoint, peer to peer

and mesh.

o Number of channels- 16 direct sequence channels

Easy to use

Communication method

Data which is transferred between the PC and the local controller is decoded by PC and Arduino

using the following manner. A bit stream consists of a letter and value. The letter indicates

specific parameter. For example t 29 .00 means temperature value is 29 ˚C. Both Aurdino and

PC can decode this data.

46 | P a g e

6. MONITORING INTERFACE

6.1. Overview

This was developed by using NI LabVIEW. LabVIEW (Laboratory Virtual Instrument

Engineering Workbench) is a graphical programming language that uses icons instead of lines

of text to create applications. In contrast to text-based programming languages, where

instructions determine the order of program execution, LabVIEW uses dataflow programming,

where the flow of data through the nodes on the block diagram determines the execution order

of the VIs and functions. VIs, or virtual instruments, are LabVIEW programs that imitate

physical instruments. [14]

a- this pane displays

current value of aerial parameters humidity, temperature, solar radiation and wind

speed.

b- hourly average values of temperature and humidity.

Figure 6-1: Monitoring Interface

47 | P a g e

c- predicted temperature and humidity values for current hour

d- calculated irrigation volume and compensation

e- field parameters evapo-transoiration and soil moisture content

f- soil moisture content graph at update rate of 1s. also indicates the maximum and minimum

soil moisture content.

g- remote irrigation and monitoring the status of the field. Required volume of water can be

irrigated and given soil moisture target can be achieved. Indications of both irrigation and

diffusion.

h- indication of communication of data reading, data logging and time synchronization.

i- Stop monitoring

6.2. Data logging capability

By reading at each and every minute data is logged in a excel file. Year, month, date, hour,

minute, temperature, humidity, wind speed, solar radiation, soil moisture content and evapo-

transpiration are logged.

And also irrigation time, amount of water, compensation or irrigation time, amount of water

are logged in another excel file for automatic irrigation and manual irrigation respectively

48 | P a g e

Figure 7-1: Mechanical Structure Design Overall View

7. MECHANICAL STRUCTURE

7.1. Design Requirements

Main objective of designing and developing the prototype was to provide an environment

which is quite similar to an actual green roof environment, for testing whether the functions

of the controller were working correctly and also to check whether the control algorithm was

able to fulfill its required tasks. In addition, as a project objective, it was decided to develop

the prototype in such a manner that it can be used for testing purposes related to the areas of

study, mentioned below.

Studying the effects on green roof evapotranspiration by changing the following

parameters (Finding crop coefficient)

Plant type

Vegetation cover

Soil texture

Roof inclination angle

Testing different kinds of irrigation control systems (Standard timer-based irrigation,

Soil moisture-based “add-on” system, Soil moisture based on-demand system,

Evapotranspiration-based system).

Acquiring actual data to predict ETo for a particular green roof.

In order to meet these requirements, following design was proposed.

7.2. Design

49 | P a g e

In the below figures, all the dimensions are in centimeters.

Figure 7-2: Mechanical Structure Isometric View

Figure 7-3: Mechanical Structure Plan View

Figure 7-4: Mechanical Structure_ Side View

50 | P a g e

Parts of the above figure and their design procedures are described below.

A- Anemometer and Temperature & humidity sensor (with the Stevenson shield [15])

location

B- Local controller

C- Plantation with soil substrates

It was decided to use the standard green roof arrangement [16] for design

purposes

Soil depth: Approximately. 12 cm deep. (Including all the substrates)

Saturated Weight: Approximately 146 kg/m2 saturated and vegetated.

Dry Weight: Approximately 100 kg/m2

D- Trays for plantation (grass)

Trays were designed to match the weight and dimensions of the substrates.

Tray dimensions: h-15cm, w-50cm, l-50cm

Material: Zinc coated metal sheets with 1mm thickness

In order to do testing regarding the irrigation requirement of plants by changing

plant cover, plant type and soil texture, it was decided to use 4 trays.

(Only one tray to be implemented by us, and provision for 3 other trays to be

kept for future testing purposes)

E- Sprinkler system

Sprinkler system was designed with micro sprinklers and they were located in

such a manner that they provide uniform distribution of water over the grass in

the tray.

F- Angle adjustable frame for carrying grass trays

As the rooftop gardens can be implemented either on horizontal or on declined

(30 degree) roofs, test bed also needed to be implemented with the capability of

varying the angle.

G- Drainage water collecting tray

In actual green roof environments, no drainage should happen. But as our task

was to develop a laboratory test bed, a tray to be used for removing drainage

water during a test was also added to the system.

51 | P a g e

H- Controlled water supply to the sprinklers

I- Solenoid valve and flow sensor

J- Water inlet to the system

K- Frame of the structure

This consists of the legs and all the other remaining parts other than what are

mentioned above. It was decided to build this with 1.5 inch L angles and 1.5

inch square hollow beams.

A, B, C, and J parts are discussed separately.

After designing the entire mechanical structure using SolidWorks, a stress analysis was

conducted to check whether the design could withstand the required loading. After verifying

that the model could meet the particular mechanical requirements, construction was done.

7.3. Construction

During the construction, it was tried hard to produce the exact output which had been

designed. But in fact there were some amendments and fabrications to be done in order to

overcome arisen problems during the physical construction. However, it was possible to

come up with a mechanical structure quite similar to the design which could meet the design

requirements.

Trays with plantation

Substrates used (From top to bottom):

Australian grass layer (1.5 cm)

Sandy loam soil layer (5 cm) – Effective root zone comes within this

Coir layer (3 cm) -

Clay brick pieces layer (2.5 cm)

Figure 7-5: Trays with Plantation

52 | P a g e

Sprinkler system

Distance between 2 adjacent sprinklers: 5 cm

Distance between the 2 sprinkler series: 18 cm

Controller

Figure 7-7: Constructed Controller

Figure 7-6: Sprinkler System

53 | P a g e

Overall structure

Parts of the above figure and their design procedures are described below.

L- Anemometer

M- Temperature and humidity sensor (with the Stevenson shield [15])

N- Local controller

O- Plantation with soil substrates

It was decided to use the standard green roof arrangement [16] for design

purposes

Soil depth: Approximately. 12 cm deep. (Including all the substrates)

Saturated Weight: Approximately 146 kg/m2 saturated and vegetated.

Dry Weight: Approximately 100 kg/m2

Figure 7-8: Constructed Mechanical Structure

54 | P a g e

P- Trays for plantation (grass)

Trays were designed to match the weight and dimensions of the substrates.

Tray dimensions: h-15cm, w-50cm, l-50cm

Material: Zinc coated metal sheets with 1mm thickness

In order to do testing regarding the irrigation requirement of plants by changing

plant cover, plant type and soil texture, it was decided to use 4 trays.

(Only one tray to be implemented by us, and provision for 3 other trays to be

kept for future testing purposes)

Q- Sprinkler system

Sprinkler system was designed with micro sprinklers and they were located in

such a manner that they provide uniform distribution of water over the grass in

the tray.

R- Angle adjustable frame for carrying grass trays

As the rooftop gardens can be implemented either on horizontal or on declined

(30 degree) roofs, test bed also needed to be implemented with the capability of

varying the angle.

S- Drainage water collecting tray

In actual green roof environments, no drainage should happen. But as our task

was to develop a laboratory test bed, a tray to be used for removing drainage

water during a test was also added to the system.

T- Controlled water supply to the sprinklers

U- Solenoid valve and floor sensor

V- Water inlet to the system

W- Frame of the structure

This consists of the legs and all the other remaining parts other than what are

mentioned above. It was decided to build this with 1.5 inch L angles and 1.5

inch square hollow beams.

A, B, C, and J parts are discussed separately.

After verifying that the model could meet the particular mechanical requirements, construction

was done.

55 | P a g e

8. CONCLUSION

Design and development of field monitoring control system for green roof as described in this

report was implemented successfully. We were able to build a laboratory prototype for

implementing our algorithm of irrigation control system. Wireless communication between

PC and the local controller was well functional.

Real time field parameters and irrigation data were logged into the separate excel sheets

successfully. Remote monitoring and remote irrigation was implemented successfully. So, this

software can be used to measure the crop coefficient of different crops, optimum water level

of any crop as the availability of logging data base and remote irrigation facility. And also

even if the wireless module is not working system operates automatically through local

controller.

As the water is scarce resource there is a vast necessity of water conservation in the vegetation

field. So, there is a market for this smart irrigating control system as it conserves water while

controlling the plant operation at its optimum condition. This product has greater potential to

be commercialized. With adequate funding and technical assistance, we can produce a

superior product that can serve our nation.

x | P a g e

9. REFERENCES

[1] “Green Roof Benefits,” 2014. [Online]. Available:

http://www.greenroofs.org/index.php/about/greenroofbenefits.

[2] “Green Roof Heat Island Effect US EPA,” 22 11 2013. [Online]. Available:

http://www.epa.gov/hiri/mitigation/greenroofs.htm.

[3] B. K. Bellingham, “Method for Irrigation Scheduling Based on Soil Moisture Data Acquisition,”

2009. [Online]. Available: http://www.stevenswater.com/articles/irrigationscheduling.aspx.

[4] R. L. Snyder,S. Eching, “Penman-Monteith (hourly) Reference Evapotranspiration,” University

of California, 2008.

[5] Garry L.Grabow,Arjun Vasanth,Dan Bowman,Rodney L. Huffman,Grady L., “EVALUATION OF

EVAPOTRANSPIRATION-BASED AND SOIL-MOISTURE-BASED IRRIGATION CONTROL IN TURF,”

North Carolina State University.

[6] Allen ,Ben, “irrigating system layout,” Ortho Books, january 2006. [Online]. Available:

http://www.watertips.com/info/layout2.htm.

[7] L. a. W. D. o. FAO, “FAO- water development and management unit-information resource- ETo

calculator,” FAO, 2013. [Online]. Available: http://www.fao.org/nr/water/eto.html.

[8] Raes. D, Timmerman. A., Hermy. M and Mentens. J, “GreenRoof water balance model,”

[Online]. Available:

http://www.iupware.be/sites/default/files/Greenroof/greenroof_manual.pdf.

[9] “Temperature and humidity module,” Aosong(Guangzhou) Electronics Co.,Ltd..

[10] “Arduino - ArduinoBoardUno,” Arduino, [Online]. Available:

http://arduino.cc/en/Main/arduinoBoardUno#.Uykq3qiSySo.

[11] “Solenoid Valve Datasheet,” Aqua Tech Trading Corp. Ltd.

[12] “Water Flow sensor,” Seed Studio Works.

[13] “XBee™ Series 2 OEM RF Modules,” www.Digi.com, 2007.

[14] LabVIEW User Manual, National Instruments Corporation, 2003.

[15] “http://en.wikipedia.org/wiki/Stevenson_screen,” [Online].

[16] “http://liveroof.com/slide/liveroof-standard/,” www.liveroof.com. [Online].

xi | P a g e

10. ANNEXES and APPENDICES

Annex 1.Arduino Code

//Libraries

#include <Time.h>

#include <TimeAlarms.h>

#include <DHT.h>

#include <ET.h>

#include <LiquidCrystal.h>

#include <Average.h>

//Pin Definition

#define soilS_pwr 2

#define flow_sensor 3

#define dht11_pin 4

#define sync_led 5

#define sol_valve 6

//Objects

DHT dht;

ET et0;

LiquidCrystal lcd(12, 11, 10, 9, 8, 7);

//Parameters

double Sub_area = 2500;//cm2

double Sub_height = 5;//cm

double flowS_factor[]=-9.13969,0.51785;//ml

double soilS_factor[]=-23,0.00058;//to volume water content

double windS_Factor=0.0246;

const double tempSF[]=-2.49807,-2.56752,-2.66011,-3.14622,-

3.14622,-3.64390,-3.23881,-2.68326,-

0.48418,0.62693,2.08526,3.70563,4.23804,4.74730,4.70100,4.16860,3.03

434,1.90008,0.61535,0.18711,-0.50733,-1.47955,-1.96566,-1.98881;

const double tempTr[]=25.980, 0.00713;

const double humSF[]=

7.7951,9.0451,9.1076,13.9201,13.8993,17.4826,19.4826,9.5660,1.8368,

-9.9965,-10.7882,-16.4965,-16.4757,-16.6424,-16.8090,-16.9549,-

12.9132,-9.2049,1.7118,1.6285,4.5243,5.0451,3.6493,7.5868;

const double humTr[]=67.77,0.0225;

const double diffution_time =6;//minutes

const double minSMT = 0.30;//volume water content

const double maxSMT = 0.48;//volume water content

//varialbles

double temp,hum,wind,solar,soil_m,evapo;

float M_temp,M_hum,M_wind;

double Acc_wind;

double Acc_T,Acc_H,Acc_W;

int J,soilS_val,windS_val;

float tempHistory[60];

float humHistory[60];

float windHistory[60];

//counters

int num_pul,lcd_count,diff_count,mincount;

int pul_count=0;

//flags

boolean irrigating=false;

boolean diffusing=false;

xii | P a g e

void setup()

lcd.begin(16, 2);

Serial.begin(9600);

pinMode(sync_led,OUTPUT);

pinMode(sol_valve,OUTPUT);

pinMode(soilS_pwr,OUTPUT);

dht.setup(dht11_pin);

et0.set(6.9344,79.861243,82.5,200,1);

setSyncProvider(sync);

//Create Timed Tasks

Alarm.timerRepeat(1,wind_update);

Alarm.timerRepeat(4,soil_update);

Alarm.timerRepeat(2,lcd_update);

Alarm.timerRepeat(60,min_update);

//initialize

initialize();

checkSync();

void loop()

checkIrr();

Alarm.delay(100);

////////////////////////////////////////////////////////////////

//Synchronizing The Time//////////////////////////////////////

void serialEvent()

char command;

command=Serial.read();

if(command=='T')

unsigned long pctime;

pctime=Serial.parseInt();

if(pctime >= 1357041600)

setTime(pctime);

checkSync();

if(command=='I')

int irr_value;

irr_value=Serial.parseInt();

irr(irr_value);

if(command=='S')

int dummyval=Serial.parseInt();

terminate_irr();

time_t sync()

Serial.write("A ");

Serial.println(0);

return 0;

////////////////////////////////////////////////////////////////

//Functions for Reading

Sensors//////////////////////////////////////////////

void readArialSensors()

xiii | P a g e

Alarm.delay(dht.getMinimumSamplingPeriod());

temp = dht.getTemperature();

hum = dht.getHumidity();

double soil_moisture()

double SMC;

soilS_val=LPF(analogRead(A0),0.1,soilS_val);

SMC=((1023-soilS_val-soilS_factor[0])*soilS_factor[1])-0.03;

if(SMC<0)SMC=0;

return SMC;

double wind_speed()

windS_val=LPF(analogRead(A1),0.1,windS_val);

return windS_val*windS_Factor;

////////////////////////////////////////////////////////////////////

///

//Send Serial Data

void serialSend()

s_out('t',temp);

s_out('h',hum);

s_out('r',solar);

s_out('w',wind);

s_out('e',evapo);

////////////////////////////////////////////////////////////////////

/

//Display Data on LCD

void lcd_update()

lcd.clear();

lcd.setCursor(0,1);//Soil Moisture Condition

lcd_out("SMC",soil_m);

lcd.print(" v/v");

lcd.setCursor(14,1);//Field Condition

if(irrigating)

lcd.print("I");

else if(diffusing)

lcd.print("D");

else if(timeStatus() != timeSet)

lcd.print("SE");

else

lcd.print("OK");

lcd.setCursor(0,0);//Weather Condition

switch(lcd_count)

case 0:

lcd_out("Temp:",temp);

lcd.print(" C");

break;

case 1:

lcd_out("Hum:",hum);

lcd.print(" %");

break;

case 2:

lcd_out("SR:",solar);

xiv | P a g e

lcd.print(" MJ/m2m");

break;

case 3:

lcd_out("Wind:",wind);

lcd.print(" m/s");

break;

case 4:

lcd_out("ET0:",evapo);

lcd.print(" mm/h");

break;

lcd_count++;

if(lcd_count>4)

lcd_count=0;

////////////////////////////////////////////////////////////////////

//Calculate Day of the Year

int Cal_J()

int days[]=0,31,59,90,120,151,181,212,243,273,304,334;

int Y = year();

int M = month();

int D = day();

if ((Y % 4 == 0 && Y % 100 != 0) || Y % 400 == 0)

return days[(M-1)]+D+1;

else

return days[(M-1)]+D;

////////////////////////////////////////////////////////////////////

////

//Timed Tasks

void wind_update()

Acc_wind+=wind_speed();

void min_update()

readArialSensors();

wind=Acc_wind/60;

Acc_wind=0;

Predict_ET();

serialSend();

s_out('T',M_temp);

s_out('H',M_hum);

s_out('W',M_wind);

if(diffusing)

if(diff_count>=diffution_time)

diff_count=0;

diffusing=false;

Serial.println("Doff");

diff_count++;

tempHistory[minute()]=temp;

humHistory[minute()]=hum;

xv | P a g e

windHistory[minute()]=wind;

if(minute()==0)

hour_update();

void soil_update()

digitalWrite(soilS_pwr,HIGH);//power Up Soil Moisture Sensor

Alarm.delay(1000);

soil_m=soil_moisture(); //Measure Soil moisture

Alarm.delay(1000);

digitalWrite(soilS_pwr,LOW);

s_out('s',soil_m);

void hour_update()

M_temp=mean(tempHistory,60);

M_hum=mean(humHistory,60);

M_wind=mean(windHistory,60);

//////////////////////////////////////////////////////////////////

//Predict ET0

void Predict_ET()

J=Cal_J();

double newTemp,newHum,newEvapo,adjEvapo;

newTemp=M_temp-tempSF[hour()-1]+tempTr[1]+tempSF[hour()];

newHum =M_hum-humSF[hour()-1]+humTr[1]+humSF[hour()];

s_out('n',newTemp);

s_out('N',newHum);

evapo = et0.ET0(newTemp,newHum,M_wind,J,hour());

solar = et0.solarRad;

/////////////////////////////////////////////////////////////////

//Irrigation Actuation

void count()

pul_count++;

if(pul_count>=num_pul)

pul_count=0;

irrigating = false;

diffusing = true;

diff_count=0;

detachInterrupt(1);

Serial.println("Ioff");

Serial.println("Don");

digitalWrite(sol_valve,LOW);

void irr(double irr_vol)

if(!irrigating)

pul_count=0;

num_pul = (int)(flowS_factor[0]+flowS_factor[1]*irr_vol);

if(num_pul>0)

attachInterrupt(1, count, RISING);

irrigating = true;

Serial.println("Ion");

xvi | P a g e

digitalWrite(sol_valve,HIGH);

void terminate_irr()

if(irrigating)

pul_count=0;

irrigating = false;

detachInterrupt(1);

Serial.println("Ioff");

digitalWrite(sol_valve,LOW);

////////////////////////////////////////////////////////////////////

//Initialize

void initialize()

soil_m=1.00;

readArialSensors();

M_temp=temp;

M_hum=hum;

M_wind=wind_speed();

for(int i=0;i<60;i++)

tempHistory[i]=temp;

humHistory[i]=hum;

windHistory[i]=wind_speed();

///////////////////////////////////////////////////////////////////

//Check The synchronization

void checkSync()

if (timeStatus() == timeSet)

digitalWrite(sync_led, HIGH); // LED on if synced

Serial.println("Son");

else

digitalWrite(sync_led, LOW); // LED off if needs refresh

Serial.println("Soff");

Predict_ET();

////////////////////////////////////////////////////////////////////

//Check Irrigation

void checkIrr()

if((soil_m < minSMT)&&(!irrigating)&&(!diffusing))

double irrigationVol=(maxSMT-minSMT)*Sub_area*Sub_height;

double compensation = evapo*0.1*Sub_area*diffution_time/60.0;

s_out('I',irrigationVol);

s_out('C',compensation);

irr(irrigationVol+compensation);

Alarm.delay(100);

////////////////////////////////////////////////////////////////////

//Serial Out

void s_out(char cmd,double value)

xvii | P a g e

Serial.print(cmd);

Serial.print(" ");

Serial.println(value);

void s_out(char cmd,int value)

Serial.print(cmd);

Serial.print(" ");

Serial.println(value);

////////////////////////////////////////////////////////////////////

/

//LCD out

void lcd_out(String cmd,double value)

lcd.print(String(cmd+" "));

lcd.print(value);

////////////////////////////////////////////////////////////////////

//Low Pass Filter

int LPF(int data, float factor, float oldVal)

float newVal;

newVal = (data * (1 - factor)) + (oldVal * factor);

return (int)newVal;

///////////////////////////////////////////////////////////////////

xviii | P a g e

Annex 2.ET Library

#include "Arduino.h"

#include "ET.h"

//double L,Lm,Lz,El,zw;

double ET::ET0(double T,double RH,double uz,int J,int t)

double Gsc = 0.082; //solar constant in MJ m-2 min-1

double sigma = 0.000000000204; //Steffan-Boltzman constant in MJ m-2

h-1 K-4

double phi = (M_PI)*L/180; // Latitude in radians

double dr =1+0.033*cos(2*(M_PI)*J/365.00); //correction for

eccentricity of Earth’s orbit around the sun

double delta = 0.409*sin(2*(M_PI)*J/365-1.39); //Declination of the

sun above the celestial equator in radians

double b = 2*(M_PI)*(J-81)/364;

double Sc = 0.1645*sin(2*b)-0.1255*cos(b)-0.025*sin(2*b); //solar

time correction for wobble in Earth’s rotation

double w =(M_PI/12)*((t-0.5)+ (Lz-Lm)/15 -12 +Sc); // hour angle in

radians

double w1=w-0.5*((M_PI)/12); // hour angle ½ hour before ? in

radians

double w2=w+0.5*((M_PI)/12);// hour angle ½ hour after ? in radians

double x =(w2-w1)*sin(phi)*sin(delta)+cos(phi)*cos(delta)*(sin(w2)-

sin(w1));

double Ra = (12*60/M_PI)*Gsc*dr*x;// extraterrestrial radiation (MJ

m-2 h-1)

double beta =

asin(sin(phi)*sin(delta)+cos(phi)*cos(delta)*cos(w))*(180/(M_PI));

//solar altitude in degrees

double Rso = Ra*(0.75+0.00002*El); // clear sky total global solar

radiation at the Earth’s surface in MJ m-2 h-1

double es = 0.6108*exp(17.27*T/(T+237.3));//saturation vapor

pressure (kPa) at the mean hourly air temperature

double ea =es*RH/100;// actual vapor pressure

double e = 0.34-0.14*sqrt(ea);//apparent ‘net’ clear sky emissivity

double Rs;

if(Rso < 0)

Rs=0;

else

Rs=Rso;

double f = (1.35*Rs/Rso)-0.35; // a cloudiness function

double Rns =(1-0.23)*Rs; //net short wave radiation as a function of

measured solar radiation

double Rnl=-f*e*sigma*pow(T+273.15,4);//net long wave radiation

double Rn =Rns+Rnl; // net radiation over grass

double Bp = 101.3*pow((293-0.0065*El)/293 , 5.26);// barometric

pressure in kPa as a function of elevation (El) in meters

double lamda = 2.45;

double gamma = 0.00163*(Bp/lamda);// psychrometric constant

double u2 = uz*(4.87/log(67.8*zw - 5.42));// wind speed at 2 m

height above ground

xix | P a g e

if(u2<0.5)

u2=0.5;

double ra=208/u2 ; //aerodynamic resistance in s m-1is estimated for

a 0.12 m tall crop as a function of wind speed (u2) in m s-1

double gammast;//Modified psychrometric constant

if(Rn > 0)

gammast = gamma*(1+0.24*u2);

else

gammast = gamma*(1+0.96*u2);

double DEL = 4099*es/pow((T+237.3),2);//slope of the saturation

vapor pressure curve (kPa oC-1) at mean air temperature (T)

double G ; // soil heat flux density (MJ m-2 h-1)

if(Rn > 0)

G = 0.1*Rn;

else

G = 0.5*Rn;

double Ro ; // Radiation term contribution to ET

if(Rn > 0)

Ro = 0.408*DEL*(Rn-G)/(DEL + gamma*(1+0.24*u2));

else

Ro = 0.408*DEL*(Rn-G)/(DEL + gamma*(1+0.96*u2));

double Ao ;//aerodynamic term of the Penman-Monteith equation in mm

d-1with u2 the wind speed at 2 m height

if(Rn > 0)

Ao = ((37*gamma/(T+273))*u2*(es-ea))/(DEL +gamma*(1+0.24*u2)) ;

else

Ao = ((37*gamma/(T+273))*u2*(es-ea))/(DEL +gamma*(1+0.96*u2));

double ETos = Ro + Ao ; //Reference evapotranspiration

solarRad = Rn;

return ETos ;

void ET::set(double _L,double _Lm,double _Lz,double _El,double _zw)

L=_L;

Lm=_Lm;

Lz=_Lz;

El=_El;

zw=_zw;

xx | P a g e

Annex 3. NI LabVIEW Block Diagram

xxi | P a g e

Appendix 1. Logged Data

Table 10-1: Part of Weather and Field Data Log

Year Month Date Hour Minutes Temperature oC

Humidity %

Wind Speed (m/s)

Solar Radiation (MJ/m2min)

ET0 (mm/hr)

Soil Moisture Content (v/v)

2014 3 18 8 59 28 63 0.14 0.89 0.41 0.41

2014 3 18 9 8 28 63 2.94 1.51 0.56 0.41

2014 3 18 10 0 29 55 2.95 2.07 0.6 0.41

2014 3 18 11 5 30 52 3.18 2.39 0.77 0.38

2014 3 18 12 12 31 44 1.95 2.59 0.79 0.36

2014 3 18 13 37 30 45 1.13 2.58 0.81 0.43

2014 3 18 14 24 30 46 1.54 2.56 0.79 0.42

2014 3 18 15 12 29 48 1.71 2.36 0.72 0.39

2014 3 18 16 53 25 64 2.88 1.97 0.59 0.39

Table 10-2: Part of Irrigation Log

Year Month Date Hour Minutes Irrigated Volume(ml) Compensation(ml)

2014 3 18 9 1 2000 0

2014 3 18 11 58 0 0

2014 3 18 12 0 1000 0

2014 3 18 12 56 2250 20.22

2014 3 18 13 14 1500 0

2014 3 18 13 22 2250 19.82

2014 3 18 14 23 2250 17.97

2014 3 18 14 46 2250 17.97

2014 3 18 14 47 2250 17.97

2014 3 18 14 57 2250 17.97

2014 3 18 14 59 1000 0

2014 3 18 14 59 1000 0

2014 3 18 14 59 1000 0

2014 3 18 14 59 1000 0

2014 3 18 15 0 1000 0

2014 3 18 15 0 1000 0

2014 3 18 15 0 2250 17.97

2014 3 18 15 9 2250 14.43

2014 3 18 15 37 2250 16.32

2014 3 18 15 57 2250 14.85

2014 3 18 16 11 2250 10.53

2014 3 18 16 24 3000 0

Design and Development of a Field Monitoring Control System for a Green Roof

Documents

Transcript of Design and Development of a Field Monitoring Control System for a Green Roof