Site Loader

Master Thesis
Smart Odour Management
Master of Science in Information Engineering
Submitted by
Hitesh Kakadiya (930017) in
in co-operation with Olfasense GmbH
Supervisor at University of Applied Science, Kiel Prof. Michael Lee
Supervisor at Olfasense GmbH Mr. Mathias Lichtner

Declaration of Authorship
I hereby declare that ˆThis Master Thesis and work reported herein was composed by and originated
entirely from me.
ˆ Information derived and verbatim from the published and unpublished work of
others has been adequately acknowledged and cited in the bibliography.
ˆ This master’s thesis has not been submitted for a higher degree at any other
University or Institution
Name: Hitesh Kakadiya
Matriculation Number: 930017
Kiel, Germany

I would like to say thanks to all the people who are involved to helping me for making
this master thesis successful one.
First and foremost, I would like to take this opportunity to express deepest sense of
gratitude towards my thesis adviser Prof. Michael Lee who has been a constant source
of inspiration throughout my thesis.
I would like to thank my mentor Mr. Mathias Lichtne and Mr. Christoph Mannebeck
at Olfsense GmbH for his outstanding supervision, ideas, energy, discussion and con-
structive feedback right from the start of my master thesis till the end.
Being a newbie to all this technologies, I had really good hands on experience with these
technologies. After completing this master thesis I came to know that learning technol-
ogy is much more dierent than applying technology practically in any application.
I am also thankful to my all company sta members to help me lot of, without their
help I can’t do. Finally, I would like to thank my family members for their never-ending
support in all my venture. I also owe my sincere thanks to all the friends who directly
or indirectly helped me with their valuable suggestions.

Confidentiality Notice
The following paper contains copyright information from Olfasense GmbH. The access
to the contents of this work is possible only for Olfasense GmbH , the author, academic
and work supervisors and reviewers of this work. The reproduction of this work or
copying of the content is prohibited.

Olfasense GmbH plans to develop a solution for the dynamic control of exhaust air
cleaning systems. Permitted installations in Europe are required to fulll certain con-
ditions regarding their emitted odorant concentrations. The aim here is to avoid the
odor nuisance of residents in the environment of industrial plants. The unit in which
smell is quantied are “odour units per m ³”. To meet these requirements, industrial
companies today install various types of exhaust air purication and ltration systems
that help minimize emitted odors. As a rule, these systems operate continuously in
24-hour operation and consume not only considerable amounts of electricity but also
consumables such as lter material or chemicals. However, depending on the location of
the industrial plant, the location of neighboring residential areas, as well as the current
local meteorology, there are many periods where odor emissions have no impact on the
neighborhood, as the wind direction and wind speed smells into uncritical (undeveloped
or agricultural) zones drives. This is where the concept of the solution to be developed
comes into play. The solution to be developed consists of a software application which
generates a signal based on an application case-related conguration of frame parameters
and location information of the industrial plant, the location of the residential areas to
be protected, local weather data and a AERMOD propagation model, which determines
when the lter systems are switched o or reduced Power can be operated. For this pur-
pose, the software should send information via an interface to a PLC controller, which
can then control the respective lter system. The development of the PLC controller
is not part of the pro ject scope. Within the scope of the pro ject, only the software
application must be realized via an interface until the data is transferred to the PLC
Keywords: Odour,Air cleaning system, AERMOD air dispersion model

Abstract iv
Contents v
List of Figuresvii
List of Tablesviii
1 Introduction1 1.1 General Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
1.2 Motivation. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
1.3 Goals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
1.4 Organization of the Thesis. . . . . . . . . . . . . . . . . . . . . . . . . . .6
2 Literature Survey7 2.1 What is Air Quality Modeling?. . . . . . . . . . . . . . . . . . . . . . . .7
2.2 A Steps required to Assessment of an Air Dispersion Modelling. . . . . .9
2.3 Overview of Air Dispersion Modelling Process. . . . . . . . . . . . . . . .9
2.4 Review of terms used during AERMOD air quality model. . . . . . . . .14
2.5 Input Data Requirements. . . . . . . . . . . . . . . . . . . . . . . . . . .15
3 System Design and Architecture19 3.1 Tools and Technologies used. . . . . . . . . . . . . . . . . . . . . . . . . .19
3.2 System Development Description. . . . . . . . . . . . . . . . . . . . . . .19 3.2.1 Motivation about Technology. . . . . . . . . . . . . . . . . . . . .22
3.3 System Development Modules. . . . . . . . . . . . . . . . . . . . . . . . .22 3.3.1 Front-end
ow Angular JS. . . . . . . . . . . . . . . . . . . . . . .22
3.3.2 Back-end
ow Hapi JS. . . . . . . . . . . . . . . . . . . . . . . . .23
3.4 System Requirements and Design. . . . . . . . . . . . . . . . . . . . . .23
3.4.1 Administrator Dashboard. . . . . . . . . . . . . . . . . . . . . . .25
3.4.2 Client Dashboard. . . . . . . . . . . . . . . . . . . . . . . . . . . .27
3.5 Challenges during implementation system. . . . . . . . . . . . . . . . . .27

vi4 Proof of Concept29
4.1 Introduction of PoCs. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29
4.2 Result and Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .34
5 Process and Cost Saving Analysis37 5.1 Type of installations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37
5.1.1 Activated Carbon. . . . . . . . . . . . . . . . . . . . . . . . . . .38
5.1.2 Scrubber. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .38
5.1.3 Thermal Treatment. . . . . . . . . . . . . . . . . . . . . . . . . .39
5.1.4 Biological Treatment. . . . . . . . . . . . . . . . . . . . . . . . . .40
6 Conclusion41 6.1 Future work. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .42

List of Figures
1.1 General graphical overview of designated tasks. . . . . . . . . . . . . . .3
1.2 Double Opportunity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4
1.3 Environment Impact. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6
2.1 Flowchart of the tasks required for assessment of air dispersion model 1.10
2.2 Data
ow of AERMOD modelling system 2. . . . . . . . . . . . . . . .12
3.1 Technology and services
ow. . . . . . . . . . . . . . . . . . . . . . . . .22
3.2 User
ow. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24
3.3 Area angle. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25
4.1 Application diagram. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30
4.2 Home Smart Odour Management. . . . . . . . . . . . . . . . . . . . . . .35
4.3 User Operations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35
4.4 Input General and Odour Parameters. . . . . . . . . . . . . . . . . . . .35
4.5 Input Impact and Area Parameters. . . . . . . . . . . . . . . . . . . . . .36
4.6 Odour Control System Status. . . . . . . . . . . . . . . . . . . . . . . . .36

List of Tables
3.1 Tools and Technologies. . . . . . . . . . . . . . . . . . . . . . . . . . . . .20
5.1 Type of Installation System. . . . . . . . . . . . . . . . . . . . . . . . . .37

AERMOD American Meteorological Society/Environmental Protection Agency Regulatory Model
AMS American Meteorological Society (AMS)
EPA Environmental Protection Agency
AERMIC (AMS)/ (EPA) Regulatory Model Improvement Committee
AERMIT AERMIC Meteorological Pre-processor
AERMAP AERMIC Terrain Pre-processor
GmbH Gesellschaft mit beschrankter Haftung
PLC Programmable Logic Controller
JS JavaScript
NPM Node Package Manager
IPPC International Plant Protection Convention
CO2 Carbon dioxide
SCRAM Support Center for Regulatory Air Models
US EPA United States Environmental Protection Agency
SBL Stable Boundary Layer
CBL Convective Boundary Layer
ISCST3 Industrial Source Complex
DEM Digital Elevation Model
USGS United States Geological Survey
VTPG Vertical Potential Temperature Gradient
WSADJ Wind Speed Adjustment and Data Source Flag
IDE Application Program Interface
API Convective Boundary Layer
CSV Comma Separated Values
JSON JavaScript Ob ject Notation

xFTP File Transfer Protocol
PoC Proof of Concept
OCU Odour Control Unit
RTO Recuperative thermal oxidation

Chapter 1
1.1 General Introduction
The environmental causes and problems of industrialization have resulted in number
of incidents of air, water and land resources being impure with noxious materials and
threatening human lives and ecosystem with health related risks. Local, regional and
global ecosystem is aected by extensive use of toxic materials.
Odours and fumes are the product of many chemical manufacturing processes.Recently
number of methods have been developed for monitoring, controlling and removing odours
which were considered something unwanted present nearby chemical plants. Chemical
plants produce odours and most of the odours are not physically dangerous but they are
psychologically dangerous. It is very dicult to prove that they do not cause organic
diseases especially when the surrounding community has been induced by rumors and
misleading information 3.
Industrial pollution is now a denite part of air pollution. Chemical engineers are
recommending and concerning with the design, supervision of chemical plants should
be according to the latest methods and techniques that are available for controlling
industrial pollution 3.

Chapter 1.
Introduction 2Industrial odours are created by a dierent type of facilities,rms and organizations,
including chemical factories, waste water treatment, food processing ,smart city man-
agement,garbage burning and many more. The problem of Odours can be created by
various industrial procedures, for instance:
ˆCertain industrial production processes emit harmful odours
ˆ Few unsuitable covers are used in waste water treatment
ˆ Microbial decomposition and decay
ˆ Organic substances available in anoxic water
ˆ Old industrial production equipment’s are needed to be repaired or maintained
Industrial odours are additionally in
uenced by environmental components; for instance,
higher temperatures quite often prompt heightened odours, and the speed or direction
of the wind can convey odours in dierent locations and directions. As towns and urban
areas proceed to develop and sprawl, this is creating more of problem. facilities that used
to be genuinely detached from the all inclusive community are confronting the challenge
of working with nearer and nearer neighbors, to removing odour is becoming critically.
An odour nuisance is thought to be pollution, as it can damage or reduce quality life of
people. Without appropriate odour control, industrial odours can
oat around dierent
area or territories and cause people to complain. An organization’s inability to follow up
on complaints got will regularly prompt ob jections being taken to local Government and
the media, which can create serious damage to a brand’s notoriety. Regardless of whether
odours are eectively conned to the plant or facility itself. Lacking odour control can
leave a business in rules and regulations of EPA and IPPC permit prerequisites, which
could nish in licenses being denied, and additionally some weighty nes.
Odour is a troublesome pollutant to distinguish,detect and dene. The human nose and
our capacity to identify and sense odour is, best case scenario sub jective, as frequently,
no two individuals smell the very same thing. Our noses are anyway preferred better
designed to measure odour over any estimating device alone. Hundreds if not a huge
number of individuals have their regular daily existences disrupted by odour pollution
over the globe consistently.

Chapter 1.
Introduction 3There are numbers of odour sources which is created in this area like agricultural, indus-
trial companies and domestic activities which may create nuisance. There are sources
of oensive odours may include the following
ˆSolid-waste processing
ˆ Food Processing
ˆ Metropolitan waste water-treatment plants
ˆ Industrial Plants
ˆ Garbage burning
The basic
ow diagram1.1shows a smart odour management system as given below :
The smart odour management system gets data from meteo station server such as wind Figure 1.1:
General graphical overview of designated tasks
direction, temperature, wind speed and more. This system will generate concentrations
by using AERMOD air dispersion model then it will send signal to PLC for enabling or
disabling air treatment system

Chapter 1.
Introduction 41.2 Motivation
Odor emissions can create serious disturbance in the area of people living particularly
in heavily populated regions. A mixture of regular aictions might be identied with
exposure to odours. Much of the time, the concentrations of odour causing chemicals
were well beneath the limit for poisonous quality. The exposure to odour may bring
about unfriendly physiological and neurogenic reactions, including stress and sickness.
With increasing population, urbanization and industrialization, the odour issue has been
creating questionable extent. Urbanization without legitimate sanitation facilities is a
signicant reason for odour nuisance. Quickly growing industrialization has irritated
the issue through odorous industrial productions. Unwanted odour adds to air quality
concerns and in
uence human’s health. Odour is without a doubt the most complex of
all the air pollution problems.
Industrial companies nowadays install various types of exhaust air purication and l-
tration systems that help minimize emitted odours. These systems operate continuously
in 24-hour operation and consume not only considerable amounts of electricity but also
consumables such as lter material or chemicals.
However, depending on the location of the industrial plant, the location of neighboring
residential areas, as well as the current local meteorology, there are many periods where
odor emissions have no impact on the neighborhood, as the wind direction and wind
speed smells into uncritical (undeveloped or agricultural) zones drives. With the use of Figure 1.2:
Double Opportunity

Chapter 1.
Introduction 5this technology we can develop a system which have potential to provide two benecial
ˆReduction of C0
2 emission
ˆ Cost saving
The chemical industries install dierent type of exhaust air purication and ltration
systems which has to be installed, maintain and keep it operational to accomplish this
they have to invest more on material, chemicals and amount of electricity. This leads
to invest more money to run this exhaust air systems. With the use of smart odour
management system we can successfully get the above mentioned benets. This is where
the concept of the solution to be developed comes into play.
1.3 Goals
Air cleaning might be helpful when utilized alongside source control and ventilation,
yet it’s anything but a substitute for either technique. The utilization of air cleaners
alone can’t guarantee adequate air quality, especially where critical sources are available
and ventilation is decient. While air cleaning may help control the levels of airborne
particles including those related with allergens and, at times, gaseous pollutants in a
home, air cleaning may not diminish antagonistic health impacts from air pollutants.
The main goal of this idea is to avoid the odor nuisance of residents in the environ-
ment near by industrial plants. To accomplish this requirement industrial companies
install various types of exhaust air purication and ltration systems that help minimize
emitted odours. Industries has certain rules and regulations to be follow and the rules
implies that the emission should not cross boundary limit of resident area if it does its
leads to series of health problems for the human body such as stress and sickness. The
rules does not implies on areas where civilization does’t exist such as forest,empty land
The exhaust air purication and ltration systems operate continuously in 24-hour op-
eration and consume not only considerable amounts of electricity but also consumables
such as lter material or chemicals.

Chapter 1.
Introduction 6Figure 1.3:
Environment Impact
The goal of smart odour management system is to reduce C0
2 emission and its cost.
1.4 Organization of the Thesis
The plan of work can be outlined as listed below: ˆGeneral Introduction
ˆ Literature Review
ˆ Implementation of Smart Odour Management System
ˆ Proof of Concept
ˆ Process and Cost Saving Analysis
ˆ Conclusion and Future work
ˆ Bibliography

Chapter 2
Literature Survey
This chapter describes literature review on air dispersion model to predict or simulate
the ambient concentrations of pollutants in the atmosphere.
The united states environmental protection agency (EPA) works on this topic since last
three decades. They already have proposed numerous models to mitigate air quality
under dierent scenarios 4. Most recent proposal is AERMOD model for industrial
sources. AERMOD will become the EPA preferred regulatory model 5.
2.1 What is Air Quality Modeling?
Air Quality Modeling is an eort to mitigate the ambient concentrations of pollutants
in the atmosphere. These models are used at primary level as a quantitative tool to
corresponded cause and eect of concentration levels that founded within the area.
Moreover, They are also used to support laws and rules of protecting air quality.
It is a tool that is utilized to evaluate the air quality eect of an emission source inside a
specied domain. The aim of dispersion model is to execute mathematical approximation
of dispersion and prepare a estimating ambient pollutant concentrations at specied
Pollutants are continuously emitted from varied sources into the air. Pollution sources
could be point sources (stacks, vents) or area sources (landlls, ponds, storage piles)
or volume sources (conveyors, structures with multiple vents). The emission of the

Chapter 2.
Literature Survey 8pollutants in the atmosphere dispersed from these sources depends on various factors.
These factors can be classied into two main categories.
Based on air quality regulations in the United States, states control the polluted emis-
sions from the dierent potential sources. All sources are delegated as ma jor or minor
sources based on their measure of emissions released every year. The resulting concen-
trations of the contamination produced from a source must be evaluated for air quality
assessment at various areas around the source. There are two fundamental ways to deal
with take care of air quality problems from stacks and vents.
They are:
1. Theoretical approach
2. Experimental approach
Each approach has its own pros and cons.
Theoretical studies incorporate analytically/numerical solutions for a few suitable equa-
tions that represent the physics of the pollutant downwind concentrations. These equa-
tions can be exceptionally straightforward or very complex relying on the desired preci-
sion and the need for assessing ground level pollutant concentration. Obviously, without
correlation with real life data, the convenience of theoretical models is restricted 6.
Experimental studies involve eld studies, wind tunnel studies, and water tank studies.
Field studies are probably going to be most precise of the two methodologies, as they
include direct measurement of pollutant concentrations at strategically located receptors.
The cons of eld studies is that they are costly to conduct and regularly require long
lead times. One can’t put any control on Mother Nature for a specic condition on a
given day. Wind tunnel and water tank studies are physical re-enactments of actual
eld study experiments. Once again, researchers have discovered that it is expensive
to mitigate exact external atmospheric conditions in an ordinary wind tunnel or water
tank study. In particular, diculty is faced in fullling the states of similarities (i.e.
scaling requirements) and the restriction of instruments to measure disturbance, mixing
height, climatic steadiness, and dierent parameters 6.
There are numbers of dispersion models accessible in the Support Center for Regula-
tory Air Models (SCRAM) part of United States Environmental Protection Agency (US

Chapter 2.
Literature Survey 9EPA) Internet website for various applications relying upon the meteorological condi-
tions,source type, the kind of study and dierent elements. The US EPA has distributed
rules on picking air quality models 3.Most of air quality dispersion model use either
the Gaussian Plume Algorithm or a variety of Gaussian Plume display.
2.2 A Steps required to Assessment of an Air Dispersion Modelling
Prior to selecting an air dispersion model, the questions ought to be asked regarding
whether an air quality model is required at all or it might be out of area to emission
point which is inappropriate and does not justify a air quality assessment.
As shown in gure2.1short overview of steps which are needed in order to assessment
of air dispersion model. In gure2.1Task 2, dispersion model input, is very crucial and
most critical outlook for air dispersion model process and requires weather data from
meteo station and other resources to ensure that the assessment of air dispersion model
is attempted eectively and successfully.
2.3 Overview of Air Dispersion Modelling Process
The procedure of air dispersion modelling consists specic information in relation with
emission source and sites to be assessed. This contains :
Source information :
ˆExit temperature
ˆ Emission rate
ˆ Exit velocity
ˆ Volume
Site information :
ˆOn site building layout

Chapter 2.
Literature Survey 10Figure 2.1:
Flowchart of the tasks required for assessment of air dispersion model 1
at or terrain information
Meteorological Information : ˆTemperature
ˆ Wind direction
ˆ Wind speed
ˆ cloud cover
Receptor Information :
ˆGrid Polar

Chapter 2.
Literature Survey 11ˆ
Discrete receptors
AERMOD is a steady-state plume model. In the stable boundary layer (SBL), it acquire
the concentration distribution to be Gaussian in both condition the vertical and horizon-
tal and for convective boundary layer (CBL) the horizontal concentration distribution
is also acquire to be Gaussian but the vertical concentration distribution is assumed to
be bi-Gaussian.
AERMOD dispersion model utilizes a Gaussian and a bi-Gaussian approach as earlier
discussed 7. It creates real-time,daily, monthly and yearly concentrations of pollutants
in the atmosphere. The model has capabilities to handles dierent type of pollutant
sources in a wide variety of settings such as rural and urban and also for
at and complex
terrain. It is a refreshed rendition of the Industrial Source Complex (ISCST3) display
being proposed by the USEPA for evaluate air quality eect from chemical factory in
coming years. One of the signicant enhancements that AERMOD brings is its capacity
to portray the planetary limit layer (PBL) through both surface and prole layer scaling.
The AERMOD model contains numerous new or enhanced calculations when contrasted
with the ISCST3 show. Some of there are below
ˆCalculation dispersion in both Stable boundary layers and Convective Buoyancy
and plume rise
ˆ Calculation of turbulence,temperature and vertical proles of wind
ˆ Treatment of surface level sources, near-surface and elevated
gure2.2shows data
ow and processing information of aermod model. The model
contain two parts one main program (AERMOD) which is consists whole model and
two pre-processors which are AERMET and AERMAP. The main use of AERMET
is to calculate boundary layer parameters for
at area. The Meteorological interface
contain prole and surface variables to calculate concentrations. AERMET passes all
meteorological data to AERMOD.
Meteorological Preprocessor (AERMET)
The fundamental aim for AERMET is to utilize meteorological measurements, delegate
of the modeling area, to gure certain boundary layer parameters used to evaluate

Chapter 2.
Literature Survey 12Figure 2.2:
ow of AERMOD modelling system 2
proles of wind, turbulence, mixing height and temperature. These proles are assessed
by the AERMOD interface which is portrayed in gure2.2.
The structure and growth of the atmospheric boundary layer control by the
uxes of heat
and momentum which is depends on surface impacts. The depth of this boundary layer
and atmospheric pollutants in
uenced by surface characteristics such as albedo,Bowen
Ratio and Surface Roughness Length and as well availability of surface humidity.
AERMET has dierent kind of surface parameters such as surface friction velocity
(u*),Monin-Obukhov Length (L), surface roughness length ( Z
o),convective scaling veloc-
ity (w*) and surface heat
ux (H). AERMET also calculate convective and mechanical
mixing height Z
i c and
i m which is very important for this model. AERMET has the
stability of the PBL by the value of H if it is (Stable H 0).
Despite the fact that AERMOD is able for evaluating meteorological proles and surface
with data from estimation of mixing height, it will use as much data as the client can

Chapter 2.
Literature Survey 13accommodate characterizing the vertical structure of the atmospheric boundary layer.
PBL parameters, AERMET passes all data of measurement of temperature,wind and
turbulence as a input in a form AERMOD needs.
Meteorological Preprocessor (AERMAP)
The module AERMAP 8 utilizes gridded territory information for the calculating area
of terrain-in
uence height which is related with every receptor area. AERMAP requires
two type of information le. The main information runstream input le that dene
prole and surface data, receptor locations and other set of options. The second sort
of information expected to run AERMAP is the Digital Elevation Model (DEM) in-
formation got from the United States Geological Survey (USGS). Amid setup handling
AERMAP checks the greater part of the sources and receptors indicated to guarantee
that they exist in the domain, and accordingly, inside the territory covered by the DEM
records. In the area that a receptor is found to lie outside the area, or if the space
reaches out past the region secured by the DEM information, AERMAP produces a
deadly mistake message, and further preparing of the information is prematurely ended.
Two kinds of landscape utilized in AERMOD are examined below:
1.Simple Terrain Simple terrain is utilized where the territory heights for the surrounding zone are
not over the highest point of the stack being assessed in the air modeling. The
“Simple” terrain is divided into two classes: i) Simple Flat Terrain is utilized
where territory rises are accepted not to surpass stack base rise. If user used this
approach then this choice is utilized, at that point terrain height is thought to
be 0.0 m, and ii) Simple Elevated Terrain is utilized where territory rises surpass
stack base yet are below stack height.
2.Complex Terrain For this situation, the territory height for the surrounding zone or area, charac-
terized as anyplace inside 50 km from the stack, are over the highest point of the
stack being assessed.

Chapter 2.
Literature Survey 142.4 Review of terms used during AERMOD air quality
The following terms and equations are frequently used in AERMOD air quality model.
A short overview is given below:
Mixing Height
The mixing height is very crucial to evaluate pollutant concentration of area. The mixing
height ( Z
i) in the CBL relies on Convective and Mechanical mixing height and is thought
to be the bigger of a mechanical mixing height ( Z
i m ) and a convective mixing height
( Z
i c ). There is two boundary layer one is SBL and another one is CBL. Mechanical
mixing height ( Z
i m ) is used in both the boundary layer SBL and CBL. While, in the
SBL, the mixing height results only from shear induced or mechanical turbulence and
consequently is equivalent to Z
i m .
Convective Mixing Height ( Z
i c )
Convective mixing height is only estimate in CBL and for SBL it will be predened
value. The height of the CBL is expected to evaluate the prole of PBL variables
and to estimate concentrations of pollutant. If there is no such a instrument available
to measurement of convective boundary layer height Z
, then Z
i c is calculated with
simple one dimension energy model (Carson 1973) which is later modied by Weil and
Brower (1983) 79.
Zi c Z
i c Z
0 Z dz
= (1 + 2 A)Z
0 H t
0 P C
t (2.1)
Here = potential temperature
A is set equal to 0.2 from Deardor (1980),
t is the hour after sunrise.
H : Sensible heat
C p : Specic heat capacity of air in
P : Air density in kgm-3
Weil and Brower discovered good understanding amongst forecasts and perceptions of
Z i c , using this model.

Chapter 2.
Literature Survey 15Mechanical Mixing Height (
i m ) :
The mechanical mixing height is calculated when
in the early morning the convective mixed layer very small or small, the full depth or
height of the PBL may be manage by mechanical turbulence. AERMET calculate height
of the PBL while convective conditions as the greatest of the evaluated convective mixing
height ( Z
i c ) and the mechanical mixing height (
i m ). AERMET utilize this method to
insure that in the early morning while convective mixing height is small or very small
but mechanical mixing height may exists.
The measurement of mechanical mixing height is not available then Z
i m is calculated
by below Zilitinkevich (1972) formula 7cite18.
Zi e = 0
:4( uL=f ); (2.2)
Z i e = Equilibrium mechanical mixing height,
F = Coriolis parameter
Venkatram (1980) also provided formula when calculation of the mechanical mixed layer
height are available then they are utilize in lieu of Z
i e .
2.5 Input Data Requirements
The source information, meteorological information and site information are basic to
run any air quality model. The information gathered must be changed over to a input
as model acceptable. The input data and related methods constitute an essential piece
of the model. Aside from these information the receptor information and the air quality
information are expected to examine the eect of a source. A short exchange is given
underneath on the distinctive classes of the information expected to run AERMOD
The AERMOD interface input les is divided in ve dierent pathways. There pathways
are :
1.SOurce information (SO)
2.MEteorology information (ME)
3.REceptor information (RE)

Chapter 2.
Literature Survey 164.EVent processing (EV)
5.OUtput options (OU)
Source information highlight the managing of various sources, including volume,point
and zone source composes. A few source gatherings might be indicated in a solitary run,
with the source commitments joined for each gathering. It additionally has dierent
features like building urban sources,downwash, and hourly emission data le.
The Control information is utilized to determine the modeling scenario, and the general
control of the run modeling. Source information is utilized to use sources of air pollutant
emissions. Receptor information is utilized to decide the air quality eect at particular
areas. Meteorology information is utilized to provide the climatic states of the zone being
displayed, which will be useful while deciding the impact of air pollution in particular
area. Territory Grid information is where the client has the choice of determining gridded
landscape information. Gridded landscape information is utilized in computing dry
exhaustion in complex territory. In the Output information the client characterizes the
kind of output results important to address the needs of air quality model analyses.
In the Meteorology information the model uses a le of prole parameters and surface
parameters like including wind direction, wind speed and others parameters. These
meteorological sources of information are generated by the meteorological preprocessor
In the Terrain Grid information the client may either utilize or passed the territory grid
information in input le or may leave the choice. If user wants to use this then user is
required to pass the location of the terrain grid le.
There are two kind of le available in AERMET :
1.Surface meteorological data le
2.Prole meteorological data le
The surface meteorological information le contains a header record containing data on
the meteorological station areas, and one record for every hour of information. These
information are delimited by no less than one space between every component, i.e., the

Chapter 2.
Literature Survey 17information might be perused as free format. The information of the surface le are as
per the following:
ˆ Month (1 { 12)
ˆ Day of Month (1 { 31)
ˆ Julian Day (Day of Year) (1 { 366)
ˆ Hour of Day (1 { 24)
ˆ Heat Flux (W/m2) : H
ˆ Surface friction velocity (m/s) : u*
ˆ Convective velocity scale (m/s) : w*
ˆ Vertical potential temperature gradient above Z
i c : VTPG
ˆ Height of convectively-generated boundary layer : Z
i c
ˆ Height of mechanically-generated boundary layer (m) : Z
ˆ Monin-Obukhov Length (m) : L
ˆ Surface Roughness Length (m) : Z
ˆ Bowen Ratio : B
ˆ Albedo : r
ˆ Reference Wind Speed (m/s) : W
ˆ Reference Wind Direction (degrees) : W
ˆ Reference Height for Wind (m) : Z
r e f
ˆ Ambient Temperature (K) : temp
ˆ Reference Height for Temperature (m) : Z
t e m p
ˆ Precipitation type code (0=none, 11=liquid, 22=frozen,99=missing) : ipcode
ˆ Precipitation Amount (mm) : pamt

Chapter 2.
Literature Survey 18ˆ
Relative Humidity (%) : rh
ˆ Station or Surface Pressure (mb) : pres
ˆ Cloud Cover (tenths) : ccvr
ˆ Wind Speed Adjustment and Data Source Flag : WSADJ
The sensible heat
ux, albedo and Bowen ratio are not utilized by the AERMOD model,
but rather are passed through by AERMET for data purposes as only.
The prole meteorological information le contains each hour of data.Similarly as with
the surface information record, the information are delimited by one space between
every variable.The substance of the prole meteorological information record are as
takes following:
ˆ Month (1 { 12)
ˆ Day of Month (1 { 31)
ˆ Hour of Day (1 { 24)
ˆ Measurement height (m)
ˆ Top
ag = 1, if this is the last (highest) level for this hour, 0, otherwise
ˆ Wind direction for the current level (degrees) : W
ˆ Wind speed for the current level (m/s) : W
ˆ Temperature at the current level (K) : T
ˆ Standard deviation of the wind direction,
2 (degrees)
ˆ Standard deviation of the vertical wind speed,
w (m/s)
The Standard deviation of the wind direction and Standard deviation of the vertical
wind speed are not utilized by the AERMOD model, but rather are passed through by
AERMET for data purposes as only.

Chapter 3
System Design and Architecture This chapter is about architecture of smart odour management system and its system
design and
ow will be discussed in details.
3.1 Tools and Technologies used
In this chapter, we will examine in detail tools and technologies utilized while creating
this system. Tools which we examine are a set of essentials to build up the system.
Moreover, We will also discuss about motivation behind using technologies for this sys-
tem, and also discuss some feature of this system, so it will help to understand thus
system. The table3.1gives detailed information about tools and technology used in
this thesis.
3.2 System Development Description
This is a full-stack JavaScript application that creates very fast, robust, and maintainable
applications using Hapi JS (Back-End), AngularJS (Front-End) and MongoDB. This is
an open source JavaScript full stack pro ject which is used to create dynamic applications.
1. JavaScript
JavaScript is an interpreted programming language that is generally used to make

Chapter 4.
System Design and Architecture 20Tools(IDE) WebStorm
Database MongoDB
Front-end HTML5, CSS3,
AngularJs Back-end) Hapi Js (Node JS
Framework) Chrome add-on Developer
tools(Debugging),Postman (REST API) Table 3.1:
Tools and Technologies
intuitive Web pages. JavaScript is one of the famous front-end programming lan-
guages and it is in control of the demeanour of a web application. I have built
up my entire pro ject in JavaScript languages which is high level and interpreted
programming languages. One of the basic innovations utilized as a part of overall
web content creation.
2. MongoDB
MongoDB was created by Elito Horowitz and Dwight Merriman, who thought
development and scalability in relational DB is a signicant issue in Web applica-
Storing data in MongoDB
Not at all like a relational database, MongoDB does not utilize tables and columns
and it is based on collections of documents. Collections incorporate heaps of ca-
pacities and archives same as relational DB table and documents compromise
key-value and additionally the essential data unit in MongoDB. MongoDB gives
dynamic composition and this component can enable document to be indicated
with various structure and elds. Also, its database utilizes a document storage
and data exchange format as known as BSON. BSON gives a binary representation
of JSON-like documents. For example, there is an arrangement of connections as
it is appeared in the gure underneath, an arrangement of TV appears, and each

Chapter 4.
System Design and Architecture 21incorporates a few seasons, and every has a few scenes, likewise every scene has
loads of cast individuals. At the point when clients visit the site, they visit the
page for their want TV appear and, on that page, they can see all seasons and
scenes. Fundamentally, clients going to a page, they will recover all data identied
with that TV appear 1011.
3. Hapi JS
Hapi JS is a Node JS framework for building micro-services for applications. It is a
conguration-centric framework which provides input validation, caching, authen-
tication, and other vital amenities. It emphases on writing reusable application
logic instead of building structure 12.
4. Angular JS
AngularJS is the most well-known Single Page Application framework kept up by
Google and it settle a ton of diculties in making and creating single-page web
applications. AngularJS is utilized for Sweco BIM because of its less inquiries to
the server to download pages, it’s easy to use include and expanding the execution
of single page application 13.
5. Node JS
Node.js is a runtime framework utilized for making server-side applications and
its fame for making ongoing Web APIs and it is utilized broadly in the tech group
because of its notoriety for adaptability, security and its eortlessness to learn.
Additionally, Node.js enables engineers to compose JavaScript on the customer
side and server-side application, which implies utilizing comparative examples and
similar libraries for front-end and back-end advancement and it can be considered
as a tremendous preferred standpoint with regards to support, designer protabil-
ity and time to showcase. Designers acquainted with JavaScript syntax structure
nd Node.js simple and they can have established Node.js on Windows or Unix in-
frastructures. Node.js gives non-blocking I/O API and speed, obviously 14. One
of the primary reasons that Node.js was built to give a superior concurrency as it
is extremely challenging in a few server-side programming languages and it causes
poor execution. Then again, Node.js oers an occasion event-driven architecture
and the non-blocking I/O API that upgrades the application’s adaptability and

Chapter 4.
System Design and Architecture 223.2.1 Motivation about Technology
As discussed, each part separately what the main reasons are to use the above technolo-
gies for development. For the database, it depends on how the data structure should
be, so according to that, we could use the relational or non-relational database to create
table or collection here in this system as per the requirement I felt that MongoDB is
best to an optimal solution to this application.
3.3 System Development Modules
The gure3.1shows architecture design for back-end and front-end services. Figure 3.1:
Technology and services
3.3.1 Front-end
ow Angular JS ˆControllers are communicating with services to call API(Services)
ˆ Services are return response from API(Services)
ˆ Views are displayed result from controllers

Chapter 4.
System Design and Architecture 233.3.2 Back-end
ow Hapi JS
ˆService is accessible only from controllers and other services.
ˆ Controllers are accessible only from routes conguration les.
ˆ Models are accessible from services only.
ˆ Policies are not accessible from anywhere, they’re just middle-ware called by the
ˆ Policies conguration le to assign dierent policies to controller’s methods.
3.4 System Requirements and Design
In this section system requirements and design will be discussed.
The system has two type of roles : ˆAdmin
ˆ Client or User
Each client or User has its own meteo station which will be hosted on site plant and
it will gather all information regarding weather forecast from its station meteo. Each
client or user has its own latitude and longitude based on the data it will operate on the
air treatment system.
Meteo station data will be stored on server and it will be in CSV format. It will be
updated every 5 min to extract new data. The system will extract meteo CSV le
format data from server to deserialize data into json format to pass it in air dispersion
model. Each user can access for logs about air treatment system when its enable or
disable. User has also access to check the percentage of activation or deactivation for
air treatment system.
Produces signal toward PLC : Signal is produced when :
1.The calculated odour plume is in a certain area/angle, as shown in gure3.3, taking into account a hysteresis area at both sides of the angle, to avoid `on/o/o/. . . .’
at border of angle.

Chapter 4.
System Design and Architecture 24Figure 3.2:

Chapter 4.
System Design and Architecture 252.The end of plume reaches a dened area/distance.
Figure 3.3:
Area angle
The signal can be
E.g.: ” = deactivate/bypass/reduce air treatment
1″ = activate/un-bypass/un-reduce air treatment
x” = amount of reduction
Software also must be able to receive signal from the PLC: e.g. if air treatment is
activated again due to other reason.
The system has two type of roles :
ˆTime registration when the plume is in a dened area when the air treatment is
e.g. deactivated
ˆ Conversion of time data to cost and C O
There are two types of dashboard in this system.
3.4.1 Administrator Dashboard
For the administrator dashboard following parameters needs to allow :
General input parameters:
ˆMinimum wind speed (the threshold which will determine when to permanently
switch on the abatement system independent from the wind direction)

Chapter 4.
System Design and Architecture 26ˆ
Hysteresis angle (“tolerance angle” in which the system is not yet switched o
even if it is outside the dened monitoring area angle)
ˆ Calculation interval (2 mins 5 mins 10mins?)
ˆ Surface roughness length (m) – z0
ˆ Latitude and Longitude of Source point
ˆ FTP Credentials for Meteo station such as Username, Password and FTP URL
Odour source settings : ˆEmission Rate (g/s) – QS
ˆ Stack Height (m) – HS
ˆ Stack exit temperature (K) – TS
ˆ Exit Velocity (m/s) – VS
ˆ Stack Diameter (m) – DS
Impact area settings : ˆAngle denition for one or more(!) areas
ˆ Distance from source point
ˆ Odour threshold settings in OU/m ³(Denes at which threshold levels the mod-
elling will trigger to activate the abatement system when a monitoring area is hit
according to the modelling – e.g. 0.5 , 1, 2)
Cost saving settings :
ˆType of Material
ˆ Material costs ( e/unit)
ˆ Material consumption (unit/h)
ˆ Type of Energy

Chapter 4.
System Design and Architecture 27ˆ
Energy costs ( e/h)
ˆ Energy consumption (unit/h)
ˆ C O
2conversion rate (factor in kg
2/kWh that is multiplied with the energy
saving amount in order to calculate the C O
2Emission savings)
3.4.2 Client Dashboard
For the client dashboard following parameters needs to allow :
Actual status of air treatment:
ˆHistorical data
Cost savings overview: ˆOver certain periods
Actual meteo + historical data e.g Graph of wind direction and wind speed
Settings Overview
3.5 Challenges during implementation system
In this part, some of the challenges during pro ject and diculties will be discussed in
detail. Some of the obstacles are as follows:
1. The problem faced during system communication with meteo station :
ˆOne of the problem faced communication between meteo station and server
for getting current weather data such as wind speed, wind direction and
temperature. Data saved in meteo server as CSV le with dierent format so
need to take care of every elements of data, otherwise it causes problems with
the data, because this data need to be passed into AERMOD Model which
calculates concentration of resident area.

Chapter 4.
System Design and Architecture 28ˆ
The system has to check current weather data every 5 min, so with received
data AERMOD model can calculate concentration. It sometime causes mem-
ory issue when it call every 5 min, because system has to fetch meteo data
from meteo server and lter it in JSON format to pass it in AERMOD model.
1. The problem faced during to nd resident area :
ˆSecond problem faced during calculating concentration from source point of
chemical or any other industries, to check whether concentration or emission
has reached resident area or not, because it is somewhat dicult to nd
coordinate of residential area due to spherical earth.
1. The challenges facing to send signal to PLC :
ˆAfter system calculated concentration its needs to send signal to PLC, so it
has to interface between system and PLC.

Chapter 4
Proof of Concept
4.1 Introduction of PoCs
Proof of Concept (PoC) is a smart method to manage business concept which consider
the business ob jectives as well as technical concept. PoCs evaluate practical potential
and viability of a concept or an idea. Eventually, it is bridges between reality and
Building an eective PoC for the task is not easy with all the future consideration we
have to gauges advantages and disadvantage which would be dicult. It is not necessary
to fulll all the requirements gathered by stakeholders in PoC.
In PoC this system doesn’t have conguration with PLC, so the system is using mocking
to prove or proposed concept.
Application diagram4.1provide general description of how the application works and
communicates in between other components.
Below is the code snippet of to get weather data from meteo station, it will extract it
to JSON to store data in database and to pass them in AERMOD model. 1
//GetweatherdatafromMeteoServer 3
varc=newClient(); 5
c.on(‘ready’,function() { 6
c.get(‘/meteoFTP_rosenow/RT/current-meas.csv’,function(err,stream) {

Chapter 5.
Proof of Concept 30Figure 4.1:
Application diagram 7
if(err)throwerr; 8
stream.once(‘close’,function() {c.end(); }); 9
stream.pipe(fs.createWriteStream(‘meteoDataLatest.csv’)); 10
stream.on(‘finish’,function() { 11
if(fs.existsSync(‘meteoDataLatest.csv’)) { 12
varreader=csv.createCsvFileReader(‘meteoDataLatest.csv’, { 14
‘separator’:’,’, 15
‘quote’:'”‘, 16
‘escape’:'”‘, 17
‘comment’:”, 18
}); 19
varallEntries=newArray(); 21
reader.setColumnNames(‘date’,’time’,’period’,’tempAir’,’humidityAir’,’ sonicTempVirtual’,’pBaro’,’sonicWsX’,’sonicWsY’
,’sonicWsZ’,’sonicWsAzimuth’,’sonicWsXYZ’,’sonicGustWs’,’sonicWdAzimuth’,’ sonicGustWdAzimuth’,’sonicWdElevation’,
‘sonicGustWdElevation’,’sonicTurbX’,’sonicTurbY’,’sonicTurbZ’,’SonicCD’,’SonicQH’,’ SonicObukhov’,’sonicStatus’,
‘sonicStatusBit0′,’sonicStatusBit1′,’sonicStatusBit2′,’sonicStatusBit3’,’sonicStatusBit4 ‘,’sonicStatusBit5′,’sonicStatusBit6′,’sonicStatusBit7′,’arkustischeVirtuelleTempU’,

Chapter 5.
Proof of Concept 3126
‘ArkustischeVirtuelleTempW’,’varianz-X-Strecke’,’varianz-Y-Strecke’,’varianz-Z-Strecke’, ‘frictionVelocity’,’schubspannung’,’dynamischeTemperatur’,’vertikalerImpulsstrom’,
‘sonicSigmaWsX’,’sonicSigmaWsY’,’SonicSigmaWsZ’,’uSupply’); AERMOD requires mixing height to evaluate pollutant concentration of area. The
mixing height consists of Convective Mixing Height ( Z
i c ) and Mechanical mixing height
( Z
i m ). To calculate the mixing height we have to use formula2.1. and2.2.
Below is the code snippet of calculating mixing height in javaScript used in the system. 1
varintegrationTime=getSecondFromTime(Meteo.time); 2
parameters.B0=Math.max(g*parameters.H0/ (rho*Cp*absoluteTemperature(parameters. T)) , 0.0).toFixed(2);//buoyancyproductionofturbulence
parameters.S0= (Math.pow(parameters.frictionVelocity, 3.0)/(k* getSurfaceRoughnessLength)).toFixed(2);//shearproductionofturbulence
parameters.turbulentHeatTransfer= (absoluteTemperature(parameters.T) *Math.pow( parameters.frictionVelocity, 2.0) / (g*k*parameters.L)).toFixed(2);
//Neutralduringdaytime 6
if(parameters.L0){ 16
varc= 0.33 /parameters.L; 17
parameters.h=Math.pow( (4.0 *c* ((B_NNM98*parameters.frictionVelocity/ getCoriolis(parameters.latitude))/100000) + 1.0) / (4.0 *Math.pow(c, 2.0)) , 0.5) –
1.0 / (2.0 *c).toFixed(2);
parameters.convectiveVelocity= -9.00; 19
parameters.Zim= (calculateStableMechanicalMixingHeight(parameters.frictionVelocity, parameters.latitude)/1000000).toFixed(2);
parameters.Zic= -999; 21
parameters.VTPG= -9.00; 22
}else{ 23
//unstableduringtheday-10000){ 5
dt=integrationTimeLeft>=TIME_STEP?TIME_STEP:integrationTimeLeft; 6
integrationTimeLeft-=TIME_STEP; 7
vars.deltaPotentialTemperature+=ddeltaThetadt(vars.TLapseRateAboveABL,params.H0,vars. h,vars.dhdt) *dt;
vars.TLapseRateAboveABL=temperatureLapseRate(params.frictionVelocity,vars. convectiveVelocity,vars.h,vars.deltaPotentialTemperature,vars.dhdt,
vars.dhdt= 2.0 * 2.5 *getPotentialTemperature(params.P,params.T) *Math.pow(params. frictionVelocity,3.0) / (g*vars.TLapseRateAboveABL*Math.pow(vars.h, 2.0));
vars.h+= (vars.dhdt*dt); 13
vars.convectiveVelocity=Math.pow(params.B0*vars.h, 1.0 / 3.0); 14
returnvars; 15
} 16
} 4.2 Result and Analysis
The primary scope of this thesis is to implement emission control and cost eective web
application, it has been achieved and below are the visual screen shots.
The gure4.2shows the home page of smart odour management. This is the default
home screen for all users.
The gure4.3shows the users list and admin can create, edit and remove users from
the list. Admin can logged into users dashboard directly from the user list panel.

Chapter 5.
Proof of Concept 35Figure 4.2:
Home Smart Odour Management Figure 4.3:
User Operations Figure 4.4:
Input General and Odour Parameters

Chapter 5.
Proof of Concept 36The gure4.4on the left screen shows the general parameters settings which includes
latitude, longitude etc. and on the right screen shows the odour source parameters which
includes emission rate, stack height etc. Figure 4.5:
Input Impact and Area Parameters
The gure4.5on the left screen shows the impact parameters settings which includes
angle from are, distance to source etc. and on the right screen shows the cost saving
parameters which includes type of material, material cost, material consumption etc.
Admin has to congure General Properties, Odour source, Impact and Area Parameters
as shown in gure4.4,4.5in order to setup user or client account. Figure 4.6:
Odour Control System Status
The gure4.6shows the status of odour control system. It will be displayed on both
admin and user console in order to get the stats. AERMOD model will return result
depending on the odour emission near resident area. The status will be either ON or

Chapter 5
Process and Cost Saving Analysis
This chapter describes air control system and how much cost could save using this
5.1 Type of installations
There are dierent type of Odour Control Unit (OCU) available. It has dierent type of
characteristics and performance. These will be considered completely while choosing the
kind of OCU that will perform best given the volume, type, variability and concentration
of the odorous parts. Consideration will likewise be given to installation, maintenance
and operability requirements of the unit.especially regarding cost, accessibility of power,
accessibility of basic extra parts and easy of access. Specically, in spite of the fact that
bad air odour system will be considered, for the most part they are not favored, and
won’t be adequate in developed territories or in nearness to neighboring inhabitants 15.
The table5.1shows the dierent type of installation system and its savings. Installations System Savings
Activated Carbon AC + energy
Scrubber Energy + chemicals
Thermal Oxidation Energy
Biolter + Scrubber Energy + material/chemicals
Cold Oxidation Energy
Table 5.1:
Type of Installation System

Chapter 6.
Process and Cost Saving Analysis 385.1.1 Activated Carbon
ˆCarbon lters are generally used to reduce emission control and odours
eect from greenhouses and dierent growing activities. The surrounding
air is circled through using activated carbon lter and its came back to
the greenhouse or released outside.
ˆ With this technology we have the potential to save AC materials to be
consume less and by which we can also save the energy which would have
been consume with this extra materials.
Indication of savings using Activated Carbon :
Activated Carbon (excl. fan savings)
ˆ10.000 m ³/h | 6.000 h/year | 100 mg/m ³VOC 10 % inactivation
ˆ 10.000 m ³/h | 6.000 h/year | 50 mg/m ³VOC 20 % inactivation
Total Saving : e12.000 savings / year
With 10.000 m ³/h activated carbon, 6.000 h/year of energy consumed by the
exhaust and with 100 mg/m ³Volatile organic compounds (VOC) we can inactivate
up to 10% if we use 50 mg/m ³VOC we can inactivate upto 20% air control system.
By this installation the total saving will be upto e12.000 per year.
5.1.2 Scrubber ˆScrubbers are odour control devices that utilization a
uid (regularly
water) to catch and remove air pollutants. Through a spout or hole
a cleaning
uid is dispersed and atomized into the gas stream. This
adequately builds the size and mass of the particles, making them simpler
to gather in a consequent separation process or lter.
ˆ The scrubbing
uid at the same time neutralizes and absorbs gaseous
air pollutants. Removed
uid is regularly recouped in fog collectors and
reused by the system.It is utilized to expel airborne contaminants before
releasing a waste air stream.
Indication of savings using Scrubber :
Ventilation (Fan) – Centrifugal vs. Axial |12.000 m ³/h | P: 2000 Pa vs. 265 Pa

Chapter 6.
Process and Cost Saving Analysis 39ˆ
6000 h/year 10 % inactivation. = Total Saving : e2.000 savings/year
ˆ 6000 h/year 20 % inactivation = Total Saving : e4.000 savings/year
ˆ 8000 h/year 10 % inactivation . = Total Saving : e2.600 savings/year
ˆ 8000 h/year 20 % inactivation . = Total Saving : e5.200 savings/year
With using 12.000 m ³/h Centrifugal, 6.000 h/year of energy consumed by the
exhaust we can inactivate up to 10 % by this installation we can save up to e2.000
per year or 20 % inactivation can save up to e4.000 per year. With the use 8.000
h/year of energy consumed by the exhaust we can inactivate up to 10 % by this
installation we can save up to e2.600 per year or 20 % inactivation can save up to
e 5.200 per year.
5.1.3 Thermal Treatment ˆThermal treatment can be essentially connected to any exhaust air. it
more often requires the expansion of pre-concentration or natural gas such
as by adsorption. But this strategy is not good because it is produces
extra emissions like sulphur oxides and nitrous.
Indication of savings using Thermal Oxidation installation :
Recuperative thermal oxidation (RTO) { (excl. fan savings) – 25.000 m ³/h | 1
g/m ³VOC at inlet air stream
ˆ 4000 h/year 10 % inactivation. = Total Saving : e12.000 savings/year
ˆ 4000 h/year 20 % inactivation = Total Saving : e24.000 savings/year
ˆ 6000 h/year 10 % inactivation . = Total Saving : e18000 savings/year
ˆ 6000 h/year 20 % inactivation . = Total Saving : e36000 savings/year
With using 25.000 m ³/h Recuperative thermal oxidation (RTO), 4.000 h/year of
energy consumed by the exhaust we can inactivate up to 10 % by this installation
we can save up to e12.000 per year or 20 % inactivation can save up to e24.000
per year. With the use 6.000 h/year of energy consumed by the exhaust we can
inactivate up to 10 % by this installation we can save up to e18.000 per year or
20 % inactivation can save up to e36.000 per year.

Chapter 6.
Process and Cost Saving Analysis 405.1.4 Biological Treatment
ˆBiological treatment is a ecologically well ,environmentally friendly and
cost saving waste air treatment strategy. It just uses organisms to absorb
the pollutants from impure air streams. Its also source of pollution but
its cost eective to compare others treatment.
The benet of using this system enables control over air treatment installation system
where the system will signal to PLC for air ventilation system turn on or o. The system
has ability to get current weather data and passes it to AERMOD model to calculate

Chapter 6
The smart odour management is a web application which has a ability to control exhaust
air cleaning system. This thesis describe in dept working of odour management system
from the initial stage to calculate pollutant concentration of area till the PLC to send
signal to exhaust air cleaning system. First, the system start to collect weather data
from meteo station and it will be pass it to AERMOD model, the calculation provided
by AERMOD will be taken to decide to air cleaning system enable or disable.
The industrial plants today install various types of exhaust air purication, cleaning
and ltration systems that help minimize emitted odors. As a rule, these systems op-
erate continuously in 24-hour operation and consume not only considerable amounts of
electricity but also consumables such as lter, material or chemicals. The goal of this
thesis is to reduce the C O
2emission which is creating by industrial plants and cost of
electricity, lter, materials and chemicals.
In chapter5, describes dierent type of air control or cleaning system, which is use to
minimize emitted odors. The air control system will be signal only when it is needed
by PLC from software. Due to non-running of air control system 24 X 7 there would
be reduction of material consumption and less usage of electricity leads to cost savings.
Material use by air control system also produce emission for cleaning odours, Less the
air control system runs less material needs so less emission produced.
The aim was to achieve two ma jor key points 1) Reduce C O
2emission and 2) Cost
savings, As mentioned in chapter4PoCs of concept its been achieved and results are

Chapter 7.
Conclusion 42discussed. This software can be implemented and the solution gives the same benet for
all the industrial plants.
6.1 Future work
With proof of work the application shows the potential of saving cost and reduce the
material usage and saving of electricity. There are still some works needs to be done to
get benet.
ˆCurrently, software is not connected with PLC, so PLC needs to be integrate
with smart odour management system. Once its integrated the air control system
working or functioning.
ˆ Existing system is passed with empty properties of cloud cover, its hard to nd
cloud cover data. System can get more accurate result with cloud cover data.
ˆ It is dicult to nd angle or co-ordinate when plume goes up and fall down at
same position. This phenomena occurs in certain industries like food processing,
garbage burning etc.

1Air dispersion modelling from industrial installations guidance note. URL
20for%20web.pdf .
2Aermod: Description of model formulation. URL
scram001/7thconf/aermod/aermod_mfd.pdf .
3Industrial odor control. URL
00966665.1958.10467834 .
4Revision to the guideline on air quality models: Adoption of a pre- ferred long range transport model and other revisions. 2003. URL
revision-to-the-guideline-on-air-quality-models-adoption-of-a-preferred-long-range-transport-model .
5Availability of additional documents relevant to anticipated revisions to guideline on air quality models addressing a preferred general purpose (
and complex terrain) dispersion model and other revisions. 2003. URL
availability-of-additional-documents-relevant-to-anticipated-revisions-to-guideline-on-air-quality .
6Understanding the usepa’s aermod modeling system. URL
download/understanding-the-usepas-aermod-modeling-system_pdf .
7User’s guide for the ams/epa regulatory model aermod. URL https://www3.epa.
gov/ttn/scram/models/aermod/aermod_userguide.pdf .
8User’s guide for the aermod terrain preprocessor (aermap). URL https://nepis. .

449Roland B. Stull. An introduction to boundary layer meteorology. 1988. URL .
10Mongo db. URL .
11Manual mongodb. URL .
12URL .
13Angular.js. URL .
14Node.js. URL .
15Odour control unit. URL
publicwebcontent/documents/document/zgrf/mdq2/ ~edisp/dd_046423.pdf

Post Author: admin