A WSN Energy-aware Approach for Air Pollution Monitoring in Waste

Treatment Facility Site: A Case Study for Landﬁll Monitoring Odour

Lelio Campanile, Mauro Iacono, Roberta Lotito and Michele Mastroianni

Dipartimento di Matematica e Fisica, Universit

a Degli Studi della Campania ”L. Vanvitelli”, Italy

Keywords:

WSN, Sensor, Energy, Monitoring, Landﬁll, Odour, LoRaWAN.

Abstract:

The gaseous emissions derived from industrial plants are generally subject to a strictly program of monitoring,

both continuous or one-spot, in order to comply with the limits imposed by the permitting license. Nowadays

the problem of odour emission, and the consequently nuisance generated to the nearest receptors, has acquired

importance so that is frequently asked a speciﬁc implementation of the air pollution monitoring program. In

this paper we studied the case study of a generic landﬁll for the implementation of the odour monitoring system

and time-speciﬁc use of air pollution control technology. The off-site monitoring is based on the deployment

of electronic nose as part of a speciﬁcally built WSN system. The nodes outside the landﬁll boundary do not

act as a continuously monitoring stations but as sensors activated when speciﬁc conditions, inside and outside

the landﬁll, are achieved. The WSN is then organized on an energy-aware approach so to prolong the lifetime

of the entire system, with signiﬁcant cost-beneﬁt advancement, and produce a monitoring-structure that can

answer to speciﬁc input like threshold overshooting.

1 INTRODUCTION

Anthropogenic environmental pollution is one of the

greatest problems that is faced and is going to be

faced by the actual generation. The pollution derived

from industrial activities is related not only to the haz-

ardous and/or accidental event (like oil spill) but also

to the regular process procedures. Indeed, almost all

the human activity emits heterogeneous pollutant into

the environment: the magnitude depends on the emis-

sion rates and on the chemical composition of the

efﬂuent. In such complex scenario, in order to de-

crease the impact and restore the environmental qual-

ity, worldwide governments and environmental agen-

cies have introduce, in their permitting protocols, pro-

cedures and threshold limits based on the Best Avail-

able Techniques (BAT) that have to be respected by

industrial plant (European Commission, 2010).

Apart from this imposition, monitoring the results

obtained by developing a Pollution Monitoring Pro-

gram (PMP) for the operating time is also important.

The PMP is edited according to the industrial plant

process and speciﬁc pollutant emission rate: it is ap-

proved as part of the permitting procedure and applied

during the industrial lifetime. Usually the monitoring

process is performed at speciﬁc time and location at

the emitting source (i.e. x chemical compound, every

six months, to be measured at the stack). Otherwise,

to establish the impact at near receptors, mathemat-

ical models are commonly used: the input data for

these systems are real data gathered after a speciﬁc

event occurred. In both cases, if during the monitor-

ing there is a non-conformity between the threshold

authorised and the value measured, the plant man-

agers must perform any procedures to reduce the im-

pact. In the worst-case scenario, the plant is shutted

down until the permitting revision.

Nowadays the necessity to have a continuous

monitoring system is raised by the plant managers,

not only to cope with the administration and local

population but also to be able to perform the nec-

essary procedures ahead of time. A typical exam-

ple is the one regarding the waste treatment facilities:

these plants are not well approved by the near popula-

tion that suffers of the so called Not In My Backyard

(NIMBY) syndrome (Xu and Lin, 2020), so people

understand the necessity to complete the integrated

waste management system, and are willing to pay the

service, but they do deny the proximity of these plants

to their homes. Therefore, the manager and plant op-

erators are called to demonstrate the respect of the

imposed limits, and so their impact, even outside the

PMP time.

In this needing scenario of an efﬁcient monitoring

526

Campanile, L., Iacono, M., Lotito, R. and Mastroianni, M.

A WSN Energy-aware Approach for Air Pollution Monitoring in Waste Treatment Facility Site: A Case Study for Landﬁll Monitoring Odour.

DOI: 10.5220/0009819005260532

In Proceedings of the 5th International Conference on Internet of Things, Big Data and Security (IoTBDS 2020), pages 526-532

ISBN: 978-989-758-426-8

system, ta Wireless Sensor Network (WSN) can be

employed. A WSN is a network formed by dislocated

nodes, linked between them by wireless transmis-

sions, so as to collect data from the surrounding en-

vironment according to installed sensors (Aiswariya

Jonsi et al., 2018; Bonastre et al., 2012). Its use in

the environmental ﬁeld is quite appealing, although

some problems have been found for real applications.

Indeed, if in the ﬁeld of waste management systems

most of the works focused on the development of

a smart bins collecting system in order to avoid the

storage overtime and ﬁnd the best collecting routes

(Narendra Kumar, 2014; Hannan et al., 2015), the

use of WSN for air quality monitoring developed in

the last decade shows various results according to

the developed architecture and sensors technical char-

acteristics. Consequentially, the marriage between

WSN and the Waste Pollution Monitoring Program

(wPMP) is not always a happy-ending one. On one

hand there is the facility manager who wants a per-

fect system, with speciﬁc characteristics such as cost-

contained and fast-answer capabilities, on the other

hand, a WSN-based system has to cope with problems

like energy supply and network modulability. Hence,

the efﬁciency of a WSN-based system depends on its

capability to identify or predict a pollutant event, re-

port it to the facility manager and apply emergency

procedures on tight deadlines.

On the basis of these requirements, Low Power

Wide Area Networks (LPWAN) acquire great im-

portance as emerging approach to Internet of Things

(IoT) and WSN: its main objective is to provide en-

ergy efﬁciency, extended coverage and scalability for

end user devices such as sensors and actuators.

LPWAN seems to be a logic choice for the aim

of this work as it provides wireless connectivity and

long range transmissions using a star topology in the

sub-Ghz frequency bands, with the major advantage

of a low power consumption which should ensure a

long sensor lifespan without the need to recharge the

battery.

Among the LPWAN technologies for IoT, there is

a growing interest for Long Range Wireless Area Net-

work (LoRaWAN), since it is a long range (from 3-

5Km in urban scenarios to 15Km in rural scenarios),

low power wide area network operating in the license

free sub-GHz bands (Centenaro et al., 2016).

This work presents a proof-of-concept architec-

ture for monitoring gaseous pollutants from waste fa-

cility plants, based on a case studio, using the Lo-

RAWAN standard. The main contribution of this ar-

chitecture is an evaluation of the application of Lo-

RAWAN in such systems to obtain better network per-

formances in term of robustness, reliability and sensor

battery life.

After this introduction, the paper is divided in ﬁve

sections. Section 2, namely Related Works, analyzes

what has been done in this speciﬁc ﬁeld. Section 3

states out the motivation that drove us and a partic-

ular case study for which we describe a speciﬁc sys-

tem architecture in section IV. Finally, conclusions are

drawn in section 5. A list of the used acronyms is pro-

vided at the end of this paper.

2 RELATED WORKS

The application of the WSN technology to air mon-

itoring is widely studied but not yet fully developed.

Indeed, the ﬁeld applications of a speciﬁc system have

to face various problems like cost and energy sup-

ply (Lee et al., 2012). According to the confronted

problem, the scientiﬁc literature is wide (Idrees and

Zheng, 2020).

In terms of environmental industrial monitoring,

a ﬁrst distinction can be made between on-site and

off-site monitoring. The on-site monitoring proce-

dures are described as the monitoring protocol per-

formed inside the plant boundary: sensors are so de-

ployed near the emitting sources or around product

units. This system is used to qualify and quantify the

local concentration and to develop secondary proce-

dures according to the emission characteristics. Some

examples of this kind of system are those related

to landﬁll sites in which both non-continuous (M

elo

et al., 2019) and continuous systems have been stud-

ied (Beirne et al., 2010).

The off-site monitoring concerns the evaluation of

the impact generated by a speciﬁc source near sensi-

tive receptors, like city centers and natural environ-

ment. In this group works are included about air

quality monitoring from a smart cities point of view

(Pauchuri, 2018; Alejandro et al., 2015; Carminati

et al., 2019).

Overall, both types of monitoring systems share

some common characteristics, like the large num-

ber of nodes deployed, and the problem of prolong-

ing network lifetime due to energy consumption. On

this last perspective, some works introduced core net-

work nodes into the system with a method based on

a jointly energy-efﬁcient development (Fadi M. Al-

Turjman, 2015) or by the use of modiﬁed algorithms

based on nodes clustering (Behera et al., 2020).

A WSN Energy-aware Approach for Air Pollution Monitoring in Waste Treatment Facility Site: A Case Study for Landﬁll Monitoring Odour

527

3 WORK MOTIVATION AND

CASE STUDY SETUP

One of the newest problems that has to be handled

by the industrial facility manager is the odour impact

generated by the normally conducted process. This

is particularly true with the waste treatment facili-

ties where, from the collecting procedures to the pro-

cess itself, there is a high probability of emission of

volatile organic compounds (VOCs) and other chem-

icals, like hydrogen sulphide, with a low detection

threshold.

The odour detection threshold (DT) is the concen-

tration at which nearly 50% of the population can de-

tect the chemical odour, so that people have a sensing

answer to it. Most of the time the limit imposed by

the permitting license overcomes the DT limit: in this

case, even if the facility is consistent with the license,

it still impacts the surrounding human environment.

The actual odour monitoring, based on the chem-

ical approach, concerns the application of a compli-

ance system and the use of speciﬁc sensors.

The compliance system works on a call-after-

impact model. Firstly, when a nuisance event occurs,

the impacted receptors call the facility administration

giving information following the what-where-when

scheme, as so to rely data about what kind of odour

they smelled, where they sensed it and when. The sec-

ond step is to match the compliance with wind speed

and direction: if the match succeeded, the compliance

is accepted. When a facility receives consecutively

compliance, it has to apply all the procedures in or-

der to reduce the emission even if these procedures

have a high cost. In the worst-case scenario, the fa-

cility is closed by the local authorities until the whole

industrial process and PMP is reviewed. Obviously,

this strategy is applied after the odour event occurred

without any possibility for warning the population nor

activating air control protocols to overthrow the pol-

lutant concentrations. Since the Air Pollution Control

technologies (APC) are usually time-cost depending

and their application on large surfaces can be cost rel-

evant in term of maintenance, and their use can also

produce a large amount of waste that has to be dis-

posed later on, the use of speciﬁc-APC has to be accu-

rate and time-speciﬁc in order to be effective, efﬁcient

and cost-contained.

Another way of odour monitoring is the use

of speciﬁc sensors called electronic nose (E-nose):

nowadays these sensors are used as a continuous mon-

itoring station deployed along the facility boundary.

The sensors are calibrated, during a so-called training

procedure, on the basis of speciﬁc pollutants emitted

under control by the source (industrial emission foot-

print). When they record a concentration above an

imposed limit, the recording hour is labelled as odour

peak event. One of the problems related to the use

of E-noses is their failure in distinguishing the back-

ground concentration – deﬁned in this case as the en-

vironmental concentration not correlated to the emit-

ting source – and the concentration derived from the

facility.

The contribution of this paper consists in the pre-

sentation of a WSN system that can answer the fa-

cility management needs to have a secure and efﬁ-

cient system that can automatically activate the APC

protocol only when a speciﬁc odour event occurs and

before the pollutants can reach the near sensitive re-

ceptors. For this reason, we set up a case study for

a generic landﬁll that still accepts organic waste, so

that it produces VOC with low DT during the normal

working procedures. The landﬁll site is equipped with

a common meteorological station – normally used for

the odour compliance system – with a logging period

of 1h. The nearest city is located 3 km far from the site

at the most frequent wind direction. To reduce vari-

ables, there is no other industrial plant that can emit

the same pollution footprint as the landﬁll: nonethe-

less, local facilities or artisans impact the air quality

by producing some of the same chemical compounds.

In this situation, the odour nuisance can be produced

either from the landﬁll or from other emitting sources.

In this scenario, the WSN needs to discern the im-

pact due to the landﬁll and the impact due to other

sources, so that the APC protocols are activated only

when a conﬁrmed odour-process is performed inside

the landﬁll and before the pollution reaches the recep-

tors.

4 SYSTEM ARCHITECTURE

The entire system is composed of three layers, char-

acterized by different nodes and roles, that exchange

information with each other and have different energy

consumption proﬁles. The ﬁrst layer is composed by

the base station and three nodes, equipped with dif-

ferent sensors, continuously recording data of com-

pound X concentration at the emitting source [X]

the wind characteristics W

direction

and W

speed

, and the

compound X background concentration [X]

: for this

layer energy supply is guaranteed by the plant itself,

so that the death of a node can be only related to actual

faults.

The second layer is formed by head cluster nodes

that, apart from the normal sensing procedures, have

the role to activate clustered nodes (third level) when

needed and collect acquired data. Hence, each head

AI4EIoTs 2020 - Special Session on Artiﬁcial Intelligence for Emerging IoT Systems: Open Challenges and Novel Perspectives

528

Figure 1: General view of the system.

cluster node HC and node N obtains a concentration

of the X compound as [X]

and [X]

Figure 1 shows the network where the HCs are

dislocated around the facility so to cover the prefer-

ential wind directions; Figure 2 shows a detail.

4.1 LoRaWAN Protocol

The LoRaWAN protocol is an open protocol based on

the proprietary LoRa physical layer. The LoRaWAN

protocol is developed by LoRa Alliance and described

in (Sornin et al., 2015).

LoRaWAN networks are organized according to a

star-of-stars topology, in which we can identify three

main components:

• End Devices (ED);

• Gateways (GW);

• a Network Server (NS);

In this topology, represented in Figure 3, each ED

transmits messages to one or multiples GWs and

each GW is connected to the NS by a stable high

bandwidth link. The responsibility of ﬁltering dupli-

cate messages from the EDs, through the GWs, is in

charge to the NS.

In our architecture HCs and N correspond to Lo-

A WSN Energy-aware Approach for Air Pollution Monitoring in Waste Treatment Facility Site: A Case Study for Landﬁll Monitoring Odour

529

Figure 2: Detail view of the system.

Figure 3: LoRaWan architecture.

RaWAN EDs, while the NS is located near the Base

Station. The Gateways could be unique and also be

placed close to the BS or, in more complex scenar-

ios where more redundancy is needed, they could be

placed close to the HCs.

All the EDs in our architecture are in Class A

mode, which according to (Sornin et al., 2015) is the

default operation mode for LoRaWAN devices. In

this mode a ED transmits the packets coming from

the upper layer on the wireless channel in an asyn-

chronous way. After the transmission, the ED wait

for any command or packet returned by the NS. Us-

ing this operational mode, devices should keep the ra-

dio transceiver off as long as possible to save battery

power.

4.2 Activation Model

The condition for the activation procedure is de-

scribed as following.

Inside the landﬁll, the ﬁrst condition to activate

the network is:

[X]

≥ DT

t = 30 min

(1)

then BS activates the HC at the preferential wind di-

rection at time t. The 30 minutes default-time is based

on the deﬁnition of odour peak, namely odour-hour

during which the odour is sensed by the population:

the restrictive time is chosen as a safety guard proce-

dure.

The HC activated at the speciﬁc direction α

◦

which the winds blows to and starts sensing the con-

centration, [X]

◦

; therefore the second condition has

to be achieved. If:

[X]

◦

− [X]

≥ DT

t = 30 min

(2)

then, a further node at α

◦

direction is activated, oth-

erwise the procedure is stopped and the HC returns to

its idle state until a new alert is stated.

The new activated nodes have to fulﬁll the second

condition for t equal to the minimum amount of time

needed for the pollutant to reach the receptor:

[X]

◦

− [X]

≥ DT

t =

speed

(3)

where d is the distance between the last deployed

node and the sensitive receptor in that direction.

The procedure continues to activate nodes until

the last is reached: if the third condition is satisﬁed,

then the APC procedure at the landﬁll is activated.

In the following the activation procedure described

above in programming metalanguage:

AI4EIoTs 2020 - Special Session on Artiﬁcial Intelligence for Emerging IoT Systems: Open Challenges and Novel Perspectives

530

1 X _s : f l o a t = 0

2 X_bk : float = 0

3 t : int = 30

4 X_hc : float = 0

5 X_n_th : float = 0

6 W_sp e e d : float = 0

7 th r e sho l d : float = 10

8 di s ta n ce _ n od e : int = 0

9 t _n : int = 0

10 w i n d_ d ir e ct i o n : [ w dir ] = None

12 w i n d_ d ir e ct i o n = g e t _ w i n d_ d ir e ct i on

()

13 wi n d _s p e ed = get_ w i nd _ sp e ed ()

15 while True :

16 if X_s >= t hre s h old :

17 hc_n o d es = g et _ h c_ n od e s (

wi n d _ d i re c ti o n )

18 for hc_nod e in hc_no d e s :

19 acti v a te ( hc_ node , t =

30)

20 X_h c = g et _ se n so r _v a lu e

( H C _ node )

21 X_hc_ t h = X _ h c - X_bk

22 if X _hc_ t h >= thr e s hol d :

23 nodes = g e t _no d e s (

wi n d _ d i re c ti o n )

24 fo r nod e in no d e s :

25 di s t an c e_ n ode =

ge t _ di s tan c e ( node )

26 acti v a t e ( node , t

= d i s tan c e_n / w i nd_ s p ee d )

27 X_n =

ge t _s e ns o r_ v al u e ( nod e )

28 X_n_th = X_n -

X_bk

29 if X _ n _th >=

thr e s hol d :

ac t iv a t e _A P C_ p r o ce d u r e ()

Listing 1: Activation Procedure.

5 CONCLUSIONS AND FUTURE

WORKS

In this work we presented a proof-of-concept archi-

tecture for monitoring gaseous pollutants, speciﬁcally

for odourant compounds, in the case study of a land-

ﬁll site. The procedure of activation of each node

is based on the fulﬁllment of three conditions to be

acquired both on-site and out-site so to activate the

air pollution control system to minimize the emission.

The adopted network protocol is LoRaWAN in order

to prolong the lifetime of the entire network and to

satisfy the speciﬁc requirements. Future work will

aim to simulate the architecture here proposed by us-

ing the ns-3 simulator, because it is widely used in

WSN simulations as you can clearly see in (Cam-

panile et al., 2020), so to understand its efﬁciency in

terms of energy consumption and throughput.

ACKNOWLEDGEMENTS

This work has been partially funded by the internal

competitive funding program “VALERE: VAnviteLli

pEr la RicErca” of Universit

a degli Studi della Cam-

pania “Luigi Vanvitelli”.

REFERENCES

Aiswariya Jonsi, S., Rani, V., and Suseela, S. (2018). Chal-

lenge, technologies and components of Wireless Sen-

sor Network. n International Journal of Engineering

Research & Technology, Special Issue.

Alejandro, V., Ferrero, R., Gandino, F., Montrucchio, B.,

and Rebaudengo, M. (2015). On the design of dis-

tributed air quality monitoring systems. AIP Confer-

ence Proceedings, 1702(1):180014.

Behera, T. M., Mohapatra, S. K., Samal, U. C., Khan, M. S.,

Daneshmand, M., and Gandomi, A. H. (2020). I-SEP:

An improved routing protocol for heterogeneous wsn

for iot-based environmental monitoring. IEEE Inter-

net of Things Journal, 7(1):710–717.

Beirne, S., Kiernan, B. M., Fay, C., Foley, C., Corcoran,

B., Smeaton, A. F., and Diamond, D. (2010). Au-

tonomous greenhouse gas measurement system for

analysis of gas migration on landﬁll sites. In 2010

IEEE Sensors Applications Symposium (SAS), pages

143–148.

Bonastre, A., Capella, J., Ors, R., and Peris, M. (2012). In-

line monitoring of chemical-analysis processes using

Wireless Sensor Networks. TrAC Trends in Analytical

Chemistry, 34:111 – 125.

Campanile, L., Gribaudo, M., Iacono, M., Marulli, F., and

Mastroianni, M. (2020). Computer network simula-

tion with ns-3: A systematic literature review. Elec-

tronics, 9(2):272.

Carminati, M., Kanoun, O., Ullo, S. L., and Marcuccio,

S. (2019). Prospects of distributed Wireless Sen-

sor Networks for urban environmental monitoring.

IEEE Aerospace and Electronic Systems Magazine,

34(6):44–52.

Centenaro, M., Vangelista, L., Zanella, A., and Zorzi, M.

(2016). Long-Range Communications in Unlicensed

Bands: the Rising Stars in the IoT and Smart City Sce-

narios. IEEE Wireless Communications, 23.

European Commission (2010). Industrial emissions direc-

tive.

Fadi M. Al-Turjman, Hossam S. Hassanein, M. I. (2015).

Towards prolonged lifetime for deployed WSNs in

outdoor environment monitoring. Ad Hoc Networks,

Volume 24, Part A:172–185.

Hannan, M., Mamun, M. A. A., Hussain, A., Basri, H.,

and Begum, R. (2015). A review on technologies and

A WSN Energy-aware Approach for Air Pollution Monitoring in Waste Treatment Facility Site: A Case Study for Landﬁll Monitoring Odour

531

their usage in solid waste monitoring and management

systems: Issues and challenges. Waste Management,

43:509 – 523.

Idrees, Z. and Zheng, L. (2020). Low cost air pollution

monitoring systems: A review of protocols and en-

abling technologies. Journal of Industrial Information

Integration, 17:100123.

Lee, K., Murray, D., Goodﬁeld, D., and Anda, M. (2012).

Experiences and issues for environmental engineering

sensor network deployments. In 2012 6th IEEE Inter-

national Conference on Digital Ecosystems and Tech-

nologies (DEST), pages 1–6.

elo, . C. A., Freire, R. C. S., Santana, E. E. C., Gurj

ao,

E. C., and Bernardo, M. G. (2019). A low cost data

acquisition system based on wireless sensor networks

for environmental monitoring in landﬁlls. In 2019 4th

International Symposium on Instrumentation Systems,

Circuits and Transducers (INSCIT), pages 1–5.

Narendra Kumar, G., C. S. N. K. (2014). Efﬁcient garbage

disposal management in metropolitan cities using

VANETs. Journal of Clean Energy Technologies,

2(3):258 – 262.

Pauchuri, V.L., R. D. (2018). IoT for monitoring carbon

monoxide (CO) emissions using wireless sensor net-

works in smart cities. International Journal of Engi-

neering & Technology, 7:1045 – 1050.

Sornin, N., Luis, M., Eirich, T., Kramp, T., and Hersent,

O. (2015). LoRaWAN speciﬁcations, LoRa Alliance,

San Ramon, CA, USA.

Xu, M. and Lin, B. (2020). Exploring the “not in my

backyard” effect in the construction of waste inciner-

ation power plants - based on a survey in metropolises

of China. Environmental Impact Assessment Review,

82:106377.

LIST OF ACRONYMS

BAT. Best Available Techniques

PMP. Pollution Monitoring Program

NIMBY. Not In My Backyard

WSN. Wireless Sensor Network

wPMP. waste Pollution Monitoring Program

VOC. Volatile Organic Compound

DT. Detection Threshold

APC. Air Pollution Control

LPWAN. Low Power Wide Area Networks

LoRaWAN. Long Range Wireless Area Network

AI4EIoTs 2020 - Special Session on Artiﬁcial Intelligence for Emerging IoT Systems: Open Challenges and Novel Perspectives

532