8

Contribution of Major Factors Affecting Non-Revenue Water to Water

Supply Network in Gaza Strip, Palestine

Jaber M. A. Alkasseh

Received Feb 2018; accepted Aug 2018

Abstract: Water losses occurring in water distribution systems (WDSs) are now considered as a serious

problem, necessitating a robust and effective management strategy. Non-revenue water (NRW) data

indicate that most cities in Gaza Strip, Palestine experience high NRW. In 2015, the average percentage

of NRW in Gaza Strip was 39%. This figure resulted in major financial, supply, and pressure losses, as

well as excessive energy consumption. The estimated annual volume of NRW in 2015 was in the order of

34 million m3 which is equivalent to American $ 16 million. Furthermore, NRW is a good indicative of

water utility performance; high levels of NRW usually indicate a poorly managed water utility. This

paper investigates the reasons why NRW is so high in many cities in Gaza Strip. Results of the study

revealed that the lack of incentives for management units, the lack of awareness of citizens-users of the

water service, apparent losses, the blockade on Gaza Strip, the wars on Gaza are the main causes.

Keywords: non-revenue water, IWA standard water balance, water service providers, key performance

indicators

Water Department, Beit Lahia Municipality, Gaza Strip, Palestine,

Email corresponding author: jalkasseh@yahoo.com

9

1. INTRODUCTION

Water is the world’s most valuable elements and one of the main sources for life (Ku-Mahamud,

Abu-Bakar and Wan-Ishak, 2005). The growing pressure on water has led this resource to be con-

sidered scarce and therefore, the efficient management of water resources is a growing necessity

(Mutikanga and Sharma, 2012). With increasing global changes such as climate change, urbaniza-

tion and population growth, there is a high probability of an additional reduction in the available

water resources in the future (Tan, Huang and Cai, 2013). This could be combined by the high rate

of water infrastructure deterioration which would cause greater loss of treated and pressurized

drinking water. Besides, the impact of poorly managed urban WDSs associated with the global

change could result in extreme scarcity scenarios (Mutikanga, Sharma and Vairavamoorthy, 2012).

Nowadays, many international organizations such as International Water Management Institute

(IWMI), The International Water Association (IWA) and World Water Council (WWC) are set up

to organise and monitor global water management (Ku-Mahamud, et al., 2005). One of the most

important issues affecting water utilities, especially in urban areas in the developing countries, is

the considerable difference between the volume of water flow into the distribution system and the

volume of water billed to consumers which is called “non-revenue water” (Kingdom, Liemberger

and Marin, 2006). In the year 2000, the IWA and American Water Works Association (AWWA)

recommended water utilities and drinking water stakeholders to use the term NRW (AWWA,

2009). The expression “water loss” and “non-revenue water” are now internationally accepted, and

have replaced expression such as “Unaccounted-For Water” (UFW) (Frauendorfer, Liemberger and

Bank, 2010).

Furthermore, the most widely accepted framework for describing NRW and determination of water

loss is the IWA Water Balance, as shown in Table 1. The IWA defines NRW as the difference

between the system input volume and billed authorized consumption. NRW comprises real or

physical losses, apparent or commercial losses, and unbilled authorized consumption (AWWA,

2009, Wyatt, 2010). Water losses in a WDS comprise apparent losses and real losses (Kanakoudis

and Tsitsifli, 2012).

Apparent losses relate to water that is being consumed but not being paid for. Apparent losses

consist of four primary components, namely customer meter inaccuracy, meter reading error, un-

authorized consumption (theft, meter bypass, illegal connections, misuse of fire hydrants, etc.), and

data handling and billing errors (Karadirek, Kara, Yilmaz, Muhammetoglu and Muhammetoglu,

2012). Apparent losses, or non-physical losses, are in many cases the most expensive water losses

to occur from a system since they represent a direct loss of revenue to the water supplier (Seago,

10

Bhagwan and McKenzie, 2007). Apparent losses component of a normal well-managed system

constitute between 10% and 20% of the total water losses (Seago, et al., 2007).

Real losses are leakage from joints in water pipes, service connections, pipe bursts, pipe cracks and

overflows from storage tanks (Karadirek, et al., 2012). Real losses can be categorised to pipe sys-

tem leakage, reservoir leakage and overflow and finally leakage from valves and pumps (Tabesh,

Yekta and Burrows, 2009). High levels of real losses normally indicate a poorly managed water

utility system (Thornton, Sturm and Kunkel, 2008); sometimes make up more than 70 % of the

total water losses (Greyvenstein and Van Zyl, 2007).

Expressing NRW and its components as % of system input volume can be very misleading. Thus,

IWA now recommends several key indicators such as NRW, physical losses, and commercial

losses, all measured in L/connection/day; as for physical losses alone, IWA recommends the use of

m3/km of pipeline/day (Wyatt, 2010). IWA has neglected using the indicator NRW as a percentage

of system input to compare locations or look at trends over time because it is accurate only if the

consumption is unchanged, which is rarely the case (Wyatt, 2010).

Table 1. IWA standard water balance and terminology (Farley and Trow, 2003)

System in-

put volume

Authorized

consumption

Billed

authorized

consump-

tion

Billed metered consumption

(including water exported) Revenue wa-

ter

Billed unmetered consump-

tion

Unbilled au-

thorized con-

sumption

Unbilled metered consump-

tion

Non-revenue

water

(NRW)

Unbilled unmetered con-

sumption

Water losses

Apparent losses Unauthorized consumption

Customer metering Inaccu-racies

Real

losses

Leakage on transmission

and/or distribution mains

Leakage and overflows at util-

ity’s storage tanks

Leakage on service connections

up to point of customer meter-

ing

11

Water losses are occurring in both developed and developing countries (Gonzalez-Gomez, García-

Rubio and Guardiola, 2011) with an estimated NRW levels of 15% and 35% of the annual system

input volume, respectively (Al-Omari, 2013).

The Global Water Supply and Sanitation Assessment 2000 Report pointed out that NRW levels in

Africa, Asia, Latin America and the Caribbean, and North America are 39%, 42%, 42%, and 15%,

respectively (Alkasseh, Adlan, Abustan, Aziz and Hanif, 2013, WHO-UNICEF-WSSCC, 2000).

Moreover, the average water loss in European Union (EU) countries is about 20%, whereas several

countries have water loss levels lower than 10% such as Germany and Denmark (Puust, Kapelan,

Savic and Koppel, 2010). A pipe network with NRW less than 15% is supposed to be in good

condition. If the value of NRW is greater than 30% the network needs immediate inspection

(Alkasseh, Adlan and Abustan, 2016). The Word Bank estimates the world wide NRW volume to

be 48.6 billion m3/year and the real losses volume (40%) occurring in the developing countries is

sufficient to supply approximately 200 million people. Furthermore, the Word Bank estimates the

monetary value of the global annual NRW volume to be US $ 14.6 billion per year (Kanakoudis

and Tsitsifli, 2012).

In the context of WDS operation and management, the sectorisation of large networks (division in

district metered areas or DMAs) can evaluate leakage level in each DMA, allowing leakage location

activities to be directed to the worst parts of the system, thus increasing their efficiency (Alkasseh,

Adlan, Abustan and Hanif, 2015). The DMA technique is first applied in the United Kingdom in

the early 1980s (Jankovic-Nisoic, Maksimovico, Butler and Graham, 2004) and has been employed

by many water utility companies worldwide (Gomes, Marques and Sousa, 2012). A DMA is de-

fined as a discrete area of distribution caused by the closure of valves, in which the quantity of

water entering and leaving the area is metered (Mounce, Boxall and Machell, 2010). It is supplied

via a single source, having approximately equal pressure levels across its population of pipes, and

with night flows regularly monitored (Fragiadakis, Christodoulou and Vamvatsikos, 2013). A per-

manently monitored DMA is considered to be the most effective tool for reducing the duration of

unreported leakage (Strum and Thorton, 2005). The introduction of DMAs and pressure manage-

ment areas (PMAs) can achieve significant reduction in real losses and frequency of bursts

(Fantozzi, Calza and Lambert, 2009).

To estimate real losses, minimum night flow (MNF) can be an indicator of distribution leakage and

consumer wastage (Johnson, Ratnayaka and Brandt, 2009). MNF is the measured flow into a con-

trolled district metered area (DMA) of a network during the period with minimum demand, i.e.,

12

between 1:00 am and 4:00 am (Johnson, et al., 2009). MNF is a commonly used method in evalu-

ating leakage levels in a water network (Loureiro, Alegre, Coelho and Borba, 2012). Generally,

MNF occurs during the early morning hours, from midnight to 4:00 am. During these hours, most

users are not consuming water; thus, water demands can be easily estimated (Cheung, Girol, Abe

and Propato, 2010). During MNF period, which usually occurs between 2:00 am and 4:00 am,

legitimate customer use is normally at a minimum, network pressures are high, and leakage is at its

maximum percentage of the total inflow into the DMA (Adlan, Alkasseh, Abustan and Hanif,

2013). Therefore, a significant increase in MNF (usually between 1:00 am and 5:00 am) can be a

reliable indicator of a system anomaly, often produced by a burst in the network (Garcia, Cabrera,

Garcia-Serra, Arregui, Almandoz, Maksimović, Butler and Memon, 2003).

A large proportion of water loss in distribution networks is common in many Asian cities, averaging

35% in the region’s cities and even reaching much higher levels (Frauendorfer, et al., 2010). In

Gaza Strip, NRW is excessive in many cities. According to the Water Sector Regulatory Council

(WSRC), the NRW in 2015 was 39% (WSRC, 2017). High levels of NRW usually indicate a poorly

managed water utility and growing pressure on water has led this resource to be considered scarce

and must therefore be managed efficiently. Considering this gap, the present study attempted to

determine the contributions of major factors that affect NRW to a water supply network in Gaza

Strip, Palestine,

2. The Study Area

The study area is situated in Gaza Strip, Palestine which is consisted of 25 cities (Figure 1). In

Gaza Strip, the groundwater is considered to be the main water source that supplies the residents

for different purposes and Gaza coastal aquifer is limited. Gaza Strip is a part of the Palestinian

coastal plain located in arid and semi-arid regions which is bordered by Egypt from the south, the

Green Line from the North, Nagev desert from the East and the Mediterranean Sea from the West

(Mayla and Amr, 2010). It is one of the most densely populated areas in the world (4138 people

per km2) based on the Municipal Development and Lending Fund (MDLF, 2009). According to

Palestinian Central Bureau of Statistics (PCBS), the estimated population in Gaza Strip at the end

of 2016 is 1.91 million (Murrar, Tamim and Samhan, 2017). Because of the high population growth

rate (~4%) (Weinthal, Vengosh, Marei, Gutierrez and Kloppmann, 2005), the population is ex-

pected to reach more than 2.6 Million inhabitants by year 2025 according to the Coastal Munici-

palities Water Utility (CMWU) and PCBS (CMWU, 2011). CMWU is the service provider for all

water and wastewater services in the Gaza strip. CMWU coordinates its activities with Palestinian

Water Authority (PWA) as a regulator for the water and wastewater sector (CMWU, 2014).

13

Figure 1. Gaza Strip Service Providers (WSRC, 2017)

The situation of the water quality and quantity in Gaza is already grim (Amr and Yassin, 2008,

UNDP, 2010). The quantities of water abstracted to meet the demand of an increased population

have put the aquifer under stress already since the late nineties (Nembrini, 2010). After Kuwait, the

Gaza Strip is the next most‘‘water-poor’’ region in the world, with 52 m3 available per person each

year (Weinthal, et al., 2005). Domestic consumption per capita was between 70 to 85 liter per

capita/ day (CMWU, 2011, UN, 2012). In addition, the aquifer has been depleted and contaminated

by over extraction and by sewage and seawater infiltration (Aiash and Mogheir, 2017, CMWU,

2011). As a consequence, , more than 90% of the water is unfit for drinking as stated by an Amnesty

International Report of 2009 (Amnesty International, 2009, CMWU, 2015) and due to high levels

14

of salinity, most of the ground water is not suitable for both domestic and agricultural consumptions

(Mayla and Amr, 2010). The concentration of chemical pollutants, including the chloride ion Cl-

and nitrate ion NO3- has exceeded the recommended WHO standards (CMWU, 2011, Mayla and

Amr, 2010). For example, the Cl- concentration varies from less than 250mg/L in the sand dune

areas as the northern and south-western area of the Gaza Strip to about more than 10,000mg/L

where the seawater intrusion has occurred (CMWU, 2011, Mayla and Amr, 2010). The NO3- con-

centration reaches a very high range in different areas of the Gaza Strip, while the WHO standard

recommended nitrate concentration less than 50mg/L (CMWU, 2011, Mayla and Amr, 2010).

3. METHODOLOGY

In Gaza Strip, there are 25 water and wastewater service providers (SPs), 24 of them are de-

partments in the municipalities and one is considered a semi-independent utility- the Coastal Mu-

nicipal Water Utility (CMWU) providing its services to Rafah Municipality (WSRC, 2016). The

general operational information and available water resources for SPs for the year 2015 are pre-

sented in Table 1.

Based on the International Water Association Performance Indicator (PI) system, a set of key per-

formance indicators (KPIs) was adopted for Palestinian water service providers. The KPIs used

were adapted to account for local conditions (WSRC, 2016). In this study, a list of technical

indicators was introduced to analyze the performance of SPs. Hence, the technical indicators were

NRW by Volume, NRW per km of Network per Year and NRW per Connection per day.

NRW by volume, as a percentage, shows the difference between the amount of water supplied

through the water distribution system and that billed to customers based on the international water

balance. It is worth noting that the water balance was used in 2015 to calculate the percentage of

NRW. The water balance is a logical analysis based on international standards in classifying the

components of NRW and provides an excellent tool for service providers to define priorities of

their action plan to reduce NRW (WSRC, 2017).

The importance of this indicator is that NRW is an indicator of the service providers’ efforts in

maintaining the assets of the utility in general, and the network in specific, in good working condi-

tions. Also, it helps the utility plan for investment in the rehabilitation or replacement of the network

and in budget preparation. Moreover, it is a tool that it is used by the regulator to assess the perfor-

mance of the individual service provider upon a request of tariff change. It is also a good monitoring

tool for the regulator to set performance levels for operators to achieve in order to safeguard the

interest of customers, reduce operation costs, and preserve limited water resources (WSRC, 2017).

15

NRW can be Calculated as: 100% - (Total billed quantity (m3) during the assessment period / (Total

supplied water during assessment period ± difference in stored quantities in utility reservoirs) *

100%).

Table 1. The general operational information and available water resources for SPs for the year 2015 (WSRC,

2017).

The relying entirely on NRW by volume as an indicator to benchmark NRW between water utilities

or even to monitor changes in performance over time can be misleading, if factors such as network

length and the number of connections are not taken into consideration (WSRC, 2016).

Municipality Water

Data 2015

Served Popula-

tion (NO.)

Active Con-

nections (NO.)

Available Wa-

ter Resources

(MCM)

Network

Length (Km)

Average daily

per capita water

consumption

(l/c/d)

Abasan al Jadida 6,114 1,279 0.39 36 131

Abasan al Kabira 23,198 3,322 1.22 55 104

Al Bureij 41,382 3,740 1.71 50 75

Al Fukhkhari 6,420 1,008 0.24 52 82

Al Maghazi 28,221 2,650 1.48 48 86

Al Mughraqa 8,241 1,412 0.67 25 94

Al Musaddar 2,410 330 0.16 19 120

Al Qarara 20,000 2,400 1.21 120 121

An Naser 8,206 1,275 0.42 37 115

An Nuseirat 75,000 8,300 4.28 93 88

Ash Shuka 11,867 1,578 0.5 82 73

Az Zahra 3,889 1,100 0.45 19 96

Az Zawayda 15,257 2,323 1.00 87 140

Bani Suheila 39,941 4,638 1.6

103

82 Beit Hanun 50,051 4,163 3.38 120 82

Beit Lahiya 73,547 7,658 4.5 170 95

CMWU – Rafah 195,570 18,34

8

7.95 375 73

Deir Al Balah 78,329 7,258 3.9 139 80

Gaza 591,712 48,13

4

31.23 600 91

Jabalia AL.Nazlh 160,157 13,67

2

12.24 190 162

Khan Yunis 193,123 17,37

9

8.45 371 72

Khuza’a 11,524 1,266 0.52 50 84

Umm an Naser 3,773 457 0.25 10 125

Wadi as Salqa 5,300 402 0.18 44 56

Wadi Gaza 3,570 318 0.12 33 39

16

Therefore, IWA and other international organizations recommend the use of the key indicators:

NRW, physical losses, and commercial losses, all measured in L/connection/day; as for physical

losses alone, IWA recommends the use of m3/km of pipeline/day (Wyatt, 2010). NRW per km of

Network per Year allows comparing service providers of different sizes and reflects the efficiency

of the network and main conveyance pipes. Moreover, it provides more accurate, reliable and com-

parable results compared to NRW percentage since it eliminates the effect of the difference in

length between networks. Also, its results will assist the water service provider to improve plans

for future investments and repair or replace the network (WSRC, 2016, WSRC, 2017).

It can be calculated as: total NRW during the year (m3) / network length (km).

Furthermore, to eliminate the impact of density of connections in the system, NRW per connection

is used along with the NRW percentage when benchmarking the performance of water utilities

(WSRC, 2016).

NRW per Connection per day can be calculated as: total NRW (m3) during the assessment period

* 1000 / number of days * total number of served connections.

3. RESULTS AND DISCUSSIONS

According to the report of PWA about the water resources status summary in 2014 in Gaza

Strip, the total water supplied for domestic use was about 88.466 million cubic meters (mcm) and

the total water consumption was 52.1 mcm. The network distribution system efficiency was 58.9%.

The NRW was 41% (PWA, 2015). The water resources available to service providers for the year

2015 were about 88.1 mcm. The amount of water purchased from Mekorot Company in 2015 was

6.9 mcm. The total cost of consumption of Mekorot Water in the Gaza Strip in 2015 was to be US

$3.4 million. The NRW was 39% (WSRC, 2017). The estimated annual volume of NRW in 2015

was to be 34 million m3. Furthermore, the estimated monetary value of the NRW volume was to be

US $ 16 million per year (WSRC, 2017). Figure 2 provides the percentages of NRW by volume of

service providers in Gaza Strip for the year 2015.

17

Figure 2. The average percentages of NRW of water SPs in Gaza Strip for the year 2015

Figure 2 confirms that most municipalities in Gaza strip experienced high NRW. Hence, the mu-

nicipalities of Al Mughraqa, Wadi Gaza and Beit Hanun still registered the highest rates of NRW

(58%, 58%, and 56% respectively). This may be due to that the majority of the municipalities in

Gaza strip were suffered from different levels of destruction in their networks during the war 2014,

which caused major losses in water (WSRC, 2016).

On 6th July 2014 the Israeli offensive on Gaza Strip has started (PWA, 2014, UNDP, 2014a) Which

resulted in a severe humanitarian crisis. According to United Nations Education, Scientific and

Cultural Organisation (UNESCO, 2014), the scale of destruction and devastation after 50 days of

conflict in July-August 2014 is unprecedented in Gaza. Consequently, water and sanitation infra-

structures were massive damaged. Based on the field survey, the main damages in the water infra-

structure revealed are as follows: (a) 15 wells were partially damaged and 11 wells were totally

damaged; (b) 11 water reservoirs were partially damaged and 5 tanks were completely damaged;

(c) a total of 20,000 metres of water network pipes of PE, UPVC and steel ranging in size from 50

mm to 315 mm, were damaged (UNDP, 2014a, UNDP, 2014b). Based on the preliminary assess-

ment, the value of damages to the water and sanitation infrastructure was estimated at USD $

34,434,100 (CMWU, 2014, UNDP, 2014a).

In the aftermath of the Gaza war 2014, the water distribution networks are highly deteriorated due

to Israeli aggressions and limited maintenance programs. Almost all municipalities in Gaza strip

had to pump water from municipal wells and distribute water through tankers (private and public)

to people who had fled from their homes during and after the war. These quantities were not billed

and were distributed free of charge (WSRC, 2016).

58 58 56

45 43 43 43 42 42 40 40 37 37 34 34 33 30 29 28 27 26 25 22 20 18

010203040506070

Al

Mu

gh

raq

a

Wad

i G

aza

Bei

t H

an

un

Az Z

ah

ra

An

Nu

seir

at

Bei

t L

ah

iya

Jab

ali

a A

L.N

azlh

Deir

Al

Bala

h

Wad

i as

Salq

a

Al

Magh

azi

Kh

an

Yu

nis

Ash

Sh

uk

a

Gaza

Al

Bu

rei

j

CM

WU

- R

afa

h

Al

Mu

sad

da

r

Um

m a

n N

ase

r

Kh

uza’a

Ab

asa

n a

l K

ab

ira

Al

Qarara

Ab

asa

n a

l Jad

ida

Ban

i S

uh

eil

a

Az Z

aw

ayd

a

Al

Fu

kh

kh

ari

An

Nase

r

NR

W b

y v

olu

me

(%)

Water sevice providers in Gaza Strip

18

In addition, a marked improvement has been noted among service providers in the rate of NRW in

2016 compared to 2015. The WSRC (2018) reported that the percentage of NRW in 2016 was 36%.

For example, the rate went down from 58% to 41% in Wadi Gaza and from 58% to 19% in Al

Mughraqa, and from 56% to 35% in Beit Hanun. Also, a large municipality like Khan Yunis

achieved a noticeable improvement in the rate of NRW which went down from 40% in 2015 to

26% only in 2016. The considerable improvement in rates of NRW can be due to the following: (i)

the improvement in the conditions of distribution networks in terms of maintenance, follow up, and

completion of repairs of sections of the network that were damaged due to the 2014 war and its

aftermath. This is clearly the case for Beit Hanun municipality, (ii) Some municipalities improved

the meter readings and treatment of illegal connections on a wide scale, as is the case of Al

Mughraqa municipality, (iii) Khan Younis municipality installed meters and billed the residential

apartments which were not billed before or the consumption readings were largely based on esti-

mates (WSRC, 2018) .

Water losses vary among systems and can be attributed to a number of different factors. These

factors include network length, number of service connections, pressure fluctuation over the day,

pipe material, leaks, bursts, and age of the system (Gomes, Marques and Sousa, 2011). Figure 3

provides the amount of NRW during the year (m3) for every km in length of the network of service

providers in Gaza Strip for the year 2015.Consequently, the figure showed a rise in the NRW per

km in the network for many SPs starting from the highest as Jabalya Al Nazla (27,466

m3/Km/year), An Nuseirat (19,982 m3/Km/year), Gaza (19,330 m3/Km/year), Beit Hanun (15,722

m3/Km/year) and others, due to the high population density and the war 2014 at the Gaza Strip, as

well as part of it is due the non-functioning administration of distribution and network wearing

(WSRC, 2017).

19

Figure 3. NRW during the year (m3)/ network length (km) of SPs in Gaza Strip for the year 2015

58 58 56 45 43 43 43 42 42 40 40 37 37 34 34 33 30 29 28 27 26 25 22 20 18

15,435

2,145

15,722

10,916

19,982

11,519

27,466

11,631

1,763

12,416

9,146

2,239

19,330

11,511

7,280

2,757

7,357

2,993

6,139

2,7722,7773,924

2,573949

2,063

Water service providers in Gaza Strip

Non-Revenue Water by volume (%) Non-revenue water per km network per year (m3/Km/year)

20

Additionally, water loss also depends on the number of service connections. Water losses increase

significantly when house connections are not done properly (Çakmakcı, Uyak, Öztürk, Aydın,

Soyer and Akça, 2007). The international studies show that the greatest proportion of losses occurs

in service connections rather than in mains, except in networks characterized by a low density of

connections (Thornton, et al., 2008). Figure 4 provides the NRW per Connection per day (l/c/d) of

service providers in Gaza Strip for the year 2015. The municipalities of Beit Hanun (1,242 l/c/d),

Jabalya Al Nazla (1,046 l/c/d) and AlMughraqa (749 l/c/d) still registered the highest rates of

NRW.

Infrastructure and services throughout Gaza’s municipalities are in need of a major overhaul as it

has suffered from years of under investment and poor maintenance due to the blockade on Gaza

(United Nations, 2008). As stated by United Nations Development Programme (UNDP) (2010),

the blockade on the Gaza Strip was imposed in June 2007. The blockade on Gaza has led to that

municipal infrastructure was nearing collapse (MDLF, 2009). For instance, the power outages,

coupled with the severe shortage of fuel and spare parts for back-up generators, have disabled part

of the water and sanitation system. Water, sanitation networks could not be rehabilitated or main-

tained for lack of spare parts and materials. As a result, about 80% of Gaza’s water wells were

functioning only partially and the remaining were non-functional. In addition, over half of the

population of Gaza City was having access to water only several hours once a week (United

Nations, 2008).

Furthermore, the electricity problem and unreliable sources are still the reason behind the negative

impacts on the operational schemes of most of the water facilities including water wells and pump-

ing facilities, and create a lot of damages and loss of services at various sites (CMWU, 2017). In

addition, the system efficiency of water distribution networks in the case of Gaza strip is subject

to the seasonal water demand requirements and temperature conditions. Consequently, system ef-

ficiency increases during winter months and starts to decrease when it comes hotter where illegal

connections especially used for irrigation. It forms about 15% of NRW which consumes a lot of

water during hot weather and negatively affects the whole system efficiency (CMWU, 2017).

21

Figure 4. NRW per Connection per day of SPs in Gaza Strip for the year 2015

58 58 56 45 43 43 43 42 42 40 40 37 37 34 34 33 30 29 28 27 26 25 22 20 18

749

610

1,242

503

613

701

1,046

611

524

616

535

318

660

422 408 439 435

324278

380

214 239 264

134 164

Water service providers in Gaza Strip

Non-Revenue Water by volume (%) Non-revenue water per connection per day (l/c/d)

22

On average, the causes of NRW are varies between developed and developing countries. In

developed countries, leakage is usually the major component, while in developing countries

apparent losses and unbilled but authorized consumption are more important. In these countries

there are a large number of illegal connections, untrue meter readings or meter damaging. More-

over, lack of incentive for utilities management to decrease the non-revenue, corruption among

utilities management, carelessness of political level and also lack of awareness campaigns for

customers are main reasons for high non-revenue water (Murrar, et al., 2017).

The high level of NRW in Gaza and West Bank is unacceptable and the PWA has a policy to

drastically reduce NRW by 2020. In 2012, PWA proposed a strategy to reduce NRW to water

service providers. Consequently, it was designed to target the reduction of NRW in the most

cost efficient manner. The PWA has a policy to reduce NRW from 38% to 35% by 2020. How-

ever initial estimates indicate, when examining the investment required, a reduction in 10% of

total estimates of the existing NRW represents in excess of $5M/yr increased revenues, or a net

present saving of over $40M compounded over 10 years. (PWA, 2012). The effort to reduce

the NRW is to be continued in the “Strategic Plan and Action Plan for the Palestinian Water

Sector (2017 – 2022)” of PWA in order to improve the efficiency of water supply. The 31%

NRW target can be achieved way before 2022 (WSRC, 2017). The 20% NRW target can be

achieved way before 2032 (WSRC, 2016).

4. CONCLUSİON AND RECOMMENDATİONS

High levels of NRW represent huge volumes of water being lost and affect the financial ca-

pability of water utilities through lost revenues and increased operational costs. Furthermore,

NRW is a good indicative of water utility performance; high levels of NRW usually indicate a

poorly managed water utility. Significant amounts of water loss are being lost because of leak-

age in WDSs. The large volume of water leakage can also cause contaminant intrusion under

low- or negative-pressure conditions within pipes, which may lead to harmful or even serious

water quality incidents. On the other hand, financial, environmental, and social benefits can be

derived from controlling and improving management of water losses caused by leakage. Hence,

minimising water lost through leakage from water supply systems is one of the main challenges

that faced water network managers. Significant amounts of money must be invested every year

for leak detection and repairs. This investment will be balanced by the benefits resulting from

23

the use of recovered water from repaired leaks. Definitely, politicians have a great responsibility

where NRW are concerned. However, Policy-makers could contribute by helping to make the

general public aware of how important it is to reduce NRW. Finally, policy-makers could also

create independent organizations to control the activity of water utilities and these control or-

ganisms should have the power to establish and monitor the accomplishment of NRW objec-

tives.

5. ACKNOWLEDGMENTS:

The author is very grateful to the Water Sector Regulatory Council (WSRC), for making

available the required data for this study. Also, the author would like to particularly thank the

Palestinian Water Authority (PWA), Coastal Municipalities Water Utility (CMWU), for their

valuable assistance.

24

References

1. Adlan, Mohd Nordin, Jaber MA Alkasseh, Hj Ismail Abustan, and Abu Bakar Mohamad

Hanif. "Identifying the Appropriate Time Band to Determine the Minimum Night Flow:

A Case Study in Kinta Valley, Malaysia." Water Science and Technology: Water Sup-

ply 13, no. 2 (2013): 328-36.

2. Aiash, Mohammed, and Yunes Khalil Mogheir. "Comprehensive Solutions for the Wa-

ter Crisis in Gaza Strip." IUG Journal of Natural Studies 25, no. 2 (2017).

3. Al-Omari, Abbas. "A Methodology for the Breakdown of Nrw into Real and Adminis-

trative Losses." Water resources management 27, no. 7 (2013): 1913-30.

4. Alkasseh, Jaber MA, Mohd Nordin Adlan, and Ismail Abustan. Economic Level of Real

Water Losses in Water Distribution System: Using Minimum Night Flow Statistical

Model. LAP LAMBERT Academic Publishing, 2016.

5. Alkasseh, Jaber MA, Mohd Nordin Adlan, Ismail Abustan, Hamidi Abdul Aziz, and

Abu Bakar Mohamad Hanif. "Applying Minimum Night Flow to Estimate Water Loss

Using Statistical Modeling: A Case Study in Kinta Valley, Malaysia." Water resources

management 27, no. 5 (2013): 1439-55.

6. Alkasseh, JMA, MN Adlan, I Abustan, and ABM Hanif. "Achieving an Economic

Leakage Level in Kinta Valley, Malaysia." Water Utility Journal 11 (2015): 31-47.

7. Amnesty International. "Troubled Waters-Palestinian Denied Fair Access to Water." Is-

rael-Occupied Palestinian Territories, 2009.

8. Amr, SS Abu, and MM Yassin. "Microbial Contamination of the Drinking Water Distri-

bution System and Its Impact on Human Health in Khan Yunis Governorate, Gaza Strip:

Seven Years of Monitoring (2000–2006)." Public Health 122, no. 11 (2008): 1275-83.

9. AWWA. Water Audits and Loss Control Programs. 3rd Edition of Manual of Water

Supply Practices M36. Denver, USA.: American Water Works Association, 2009.

10. Çakmakcı, M., V. Uyak, İ. Öztürk, A. F. Aydın, E. Soyer, and L. Akça. "The Dimension

and Significance of Water Losses in Turkey." In Proceedings of IWA Specialist Confer-

ence on Water Loss, , 464-73. Bucharest, Romania,, 2007.

11. Cheung, P.B., G.V. Girol, N. Abe, and M. Propato. "Night Flow Analysis and Modeling

for Leakage Estimation in a Water Distribution System." Integrating Water Systems–

Maksimovic, Taylor & Francis Group (2010): 509-13.

12. CMWU. "Cmwu Annual Report on Water Status in the Gaza Strip." Summary about

Water and Wastewater Situation in Gaza Strip, 2011.

13. CMWU. "Water and Wastewater Infrastructure Damage Assessment (Gaza Strip; 7 July

- 14 August 2014)." 2014.

14. CMWU. "Water & Wastewater Sectors Contingency Plan for Conflict Times." Gaza

Strip, 2015.

15. CMWU. "Annual Report 2016." (2017).

16. Fantozzi, M., F. Calza, and A. Lambert. "Experience and Results Achieved in Introduc-

ing District Metered Areas (Dma) and Pressure Management Areas (Pma) at Enia Util-

ity (Italy). In Water Loss Specialist Conference, International Water Association, Cape

Town, South Africa. 17." 2009.

25

17. Farley, Malcolm, and Stuart Trow. Losses in Water Distribution Networks; a Practi-

tioner’s Guide to Assessment, Monitoring and Control. London, UK.: IWA Publishing,,

2003.

18. Fragiadakis, Michalis, Symeon E Christodoulou, and Dimitrios Vamvatsikos. "Reliabil-

ity Assessment of Urban Water Distribution Networks under Seismic Loads." Water re-

sources management 27, no. 10 (2013): 3739–64.

19. Frauendorfer, R., R. Liemberger, and Asian Development Bank. The Issues and Chal-

lenges of Reducing Non-Revenue Water. Asian Development Bank, Mandaluyong City,

Philippines., 2010.

20. Garcia, VJ, E. Cabrera, J. Garcia-Serra, F. Arregui, J. Almandoz, C. Maksimović, D.

Butler, and F. Memon. "Stochastic Prediction of the Minimum Night Flow Demand in a

District Metered Area." Paper presented at the Advances in Water Supply Management,

Proceedings of the CCWI '03 Conference, London, UK, 2003.

21. Gomes, R., A.S. Marques, and J. Sousa. "Estimation of the Benefits Yielded by Pressure

Management in Water Distribution Systems." Urban Water Journal 8, no. 2 (2011): 65-

77.

22. Gomes, R., A.S. Marques, and J. Sousa. "Identification of the Optimal Entry Points at

District Metered Areas and Implementation of Pressure Management." Urban Water

Journal, Taylor & Francis, DOI:10.1080/1573062X.2012.682589 (2012).

23. Gonzalez-Gomez, F., M.A. García-Rubio, and J. Guardiola. "Why Is Non-Revenue Wa-

ter So High in So Many Cities?". Water Resources Development 27, no. 02 (2011): 345-

60.

24. Greyvenstein, B., and JE Van Zyl. "An Experimental Investigation into the Pressure-

Leakage Relationship of Some Failed Water Pipes." Aqua- Journal of Water Supply:

Research and Technology 56, no. 2 (2007): 117-24.

25. Jankovic-Nisoic, B., C. Maksimovico, D. Butler, and NJD Graham. "Use of Flow Me-

ters for Managing Water Supply Networks." Journal of Water Resources Planning and

Management 130, no. 2 (2004): 171-79.

26. Johnson, Michael, Don D. Ratnayaka, and Malcom J. Brandt. Twort’s Water Supply.

6th ed. Oxford, UK: Elsevier Ltd, 2009.

27. Kanakoudis, V., and S. Tsitsifli. "Urban Water Services Public Infrastructure Projects:

Turning the High Level of the Nrw into an Attractive Financing Opportunity Using the

Pbsc Tool." Desalination and Water Treatment 39, no. 1-3 (2012): 323-35.

28. Karadirek, IE, S. Kara, G. Yilmaz, A. Muhammetoglu, and H. Muhammetoglu. "Imple-

mentation of Hydraulic Modelling for Water-Loss Reduction through Pressure Manage-

ment." Water resources management 26, no. 9 (2012): 2555–68.

29. Kingdom, Bill, Roland Liemberger, and Philippe Marin. "The Challenge of Reducing

Non-Revenue Water (Nrw) in Developing Countries, Water Supply and Sanitation Sec-

tor Board Discussion Paper Series, Paper No.8.". (2006).

30. Ku-Mahamud, Ku Ruhana, Muhamad Shahbani Abu-Bakar, and Wan Hussain Wan-

Ishak. "Modeling the Perlis Non Revenue Water." Paper presented at the Proceeding of

the First International Conference on Modeling, Simulation and Applied Optimization,

Sharjah, U.A.E., 2005.

31. Loureiro, Dália, Helena Alegre, Sérgio Teixeira Coelho, and Rodrigo Borba. "A New

Approach to Estimatinghousehold Night Consumption at Dma Level." Paper presented

26

at the Procceedings of the 7th IWA Water Loss Reduction Specialist Conference, Ma-

nila, Philippines, 2012.

32. Mayla, Yousef Salah Abu, and Salem Saleem Abu Amr. "Chemical and Microbiological

Quality of Drinking Water in Gaza Strip, Palestine." Science Vision 10, no. 3 (2010):

80-88.

33. MDLF. "Gaza Municipal Sector Damage Assessment, Draft for Discussion." This report

was endorsed by the Ministry of Local Government (MoLG) as Part of the National

Plan to Rebuild Gaza., 2009.

34. Mounce, SR, JB Boxall, and J. Machell. "Development and Verification of an Online

Artificial Intelligence System for Detection of Bursts and Other Abnormal Flows." Jour-

nal of Water Resources Planning and Management 136(3) (2010): 309–18.

35. Murrar, Abdullah, Abdelrahman Tamim, and Subhi Samhan. "The Determinants of

Non-Revenue Water & Financial Viability for the Palestinian Water Service Providers."

Journal of Water Resources and Ocean Science 6, no. 2 (2017): 35-45.

36. Mutikanga, H.E., and S.K. Sharma. "Strategic Planning for Water Loss Reduction with

Imprecise Data." Paper presented at the Procceedings of the 7th IWA Water Loss Re-

duction Specialist Conference, Manila, Philippines, 2012.

37. Mutikanga, H.E., S.K. Sharma, and K. Vairavamoorthy. "Review of Methods and Tools

for Managing Losses in Water Distribution Systems." Journal of Water Resources Plan-

ning and Management. doi:10.1061/(ASCE)WR.1943-5452.0000245 (2012).

38. Nembrini, PG. "The Gaza Strip: The State of the Water Supply after the 2008–2009

War." Cities in war, thirsty cities (2010): 1-28.

39. Puust, R., Z. Kapelan, DA Savic, and T. Koppel. "A Review of Methods for Leakage

Management in Pipe Networks." Urban Water Journal 7, no. 1 (2010): 25-45.

40. PWA. "Non-Revenue Water Reduction Strategy-Final Draft." (2012).

41. PWA. "Water Sector Damage Assessment Report." This Report has been prepared in

full coordination and cooperation with CMWU, 2014.

42. PWA. "2014 Water Resources Status Summary Report /Gaza Strip." (2015).

43. Seago, C., J. Bhagwan, and R. McKenzie. "Benchmarking Leakage from Water Reticu-

lation Systems in South Africa." Water SA 30, no. 5 (2007): 25-32.

44. Strum, R., and J. Thorton. "Proactive Leakage Management Using District Metered Ar-

eas (Dma) and Pressure Management–Is It Applicable in North America." Paper pre-

sented at the IWA Leakage 2005 conference proceedings, 2005.

45. Tabesh, M., A.H.A. Yekta, and R. Burrows. "An Integrated Model to Evaluate Losses in

Water Distribution Systems." Water Resources Management 23, no. 3 (2009): 477-92.

46. Tan, Q., GH Huang, and YP Cai. "Multi-Source Multi-Sector Sustainable Water Supply

under Multiple Uncertainties: An Inexact Fuzzy-Stochastic Quadratic Programming Ap-

proach." Water resources management 27 (2013): 451–73.

47. Thornton, Julian, Reinhard Sturm, and George Kunkel. Water Loss Control. Second Ed.

New York, USA: McGraw-Hill 2008.

48. UN. "Gaza in 2020 a Liveable Place?": A report by the United Nations Country Team in

the occupied Palestinian territory, 2012.

49. UNDP. "Early Recovery and Reconstruction Needs Assessment– One Year After." Pro-

gramme of Assistance to the Palestinian People, Gaza, 2010.

50. UNDP. "Detailed Infrastructure Damage Assessment." GAZA, 2014a.

27

51. UNDP. "Gaza Crisis - 2014 - Preliminary Assessment." Jerusalem: Prepared by the Ap-

plied Research Institute, 2014b.

52. United Nations. "The Impact of the Blockade on the Gaza Strip." Gaza Humanitarian

Situation Report: Office for the Coordination of Humanitarian Affairs, 2008.

53. Weinthal, Erika, Avner Vengosh, Amer Marei, Alexis Gutierrez, and Wolfram

Kloppmann. "The Water Crisis in the Gaza Strip: Prospects for Resolution." Groundwa-

ter 43, no. 5 (2005): 653-60.

54. WHO-UNICEF-WSSCC. "Global Water Supply and Sanitation Assessment 2000 Re-

port." New York: World Health Organization and United Nations Children’s Fund,

2000.

55. WSRC. "Performance Monitoring of Water Service Providers in Palestine for the Year

2014." (2016).

56. WSRC. "Bridge to Sustainability Water and Wastewater Service Providers in Palestine:

Facts and Prospects - 2015 Report." (2017).

57. WSRC. "Bridge to Sustainability Water and Wastewater Service Providers Performance

Monitoring Report for the Year 2016." (2018).

58. 58. Wyatt, Alan S. "Non-Revenue Water: Financial Model for Optimal Management in

Developing Countries. Rti Press Publication Mr-0018-1006." (2010).

1 Environmental Quality Authority, Gaza City, Gaza Strip, Palestine (rana.idais@yahoo.com) 2 Faculty of Engineering, The Islamic University of Gaza, P.O. Box 108, Gaza Strip, Palestine 3BZU Testing Laboratories, Birzeit University, Ramallah, Palestine

Evaluation the Impact of Using Treated Wastewater for Irrigation

on Soil Chemical Properties and Crop Productivity in Gaza Strip

R. H. Idais*1 A. R. Nassar 2 and A.R. Mughari3

Received Feb 2018; accepted Apr 2018

ABSRACT :The aim of study is to investigate the impact of using treated wastewater for irrigation

on soil chemical properties and plant productivity. An reuse pilot study was carried out in Al-Zaitoun

agricultural farm in the Gaza Strip from May to September 2011. A comparison was carried out

between the soil properties in two experimental plots; one was irrigated with the effluent from Gaza

Wastewater Treatment Plant over a period of four months, and the other was irrigated with fresh

water from agricultural well in the same period. The crop used was sorghum. Samples of fresh water

and treated wastewater were obtained and analyzed for pH, TDS, EC, Na, Cl, TKN, TP, K, BOD,

NO3, TSS, FC, Fe, Mn, Zn. Composite soil samples were taken from depth of 0-30 cm in both plots

and analyzed for the main chemical parameters. The results indicated that the level of TDS, Na, Cl,

TSS, Zn and Fe were higher in the effluent than the fresh water; it was above the recommended

Palestinian standard for dry fodder irrigated by treated wastewater. Also, irrigation with wastewater

lead to significant increase in O.M%, CEC, K, TP, Ca, Mg, Na, and Cl in soil than irrigation with

fresh water. In addition, the increases of Zn, Fe, Mn, and Pb in soil and sorghum plant irrigated

with treated wastewater were significant in comparison with the plants irrigated with fresh water.

Further, treated wastewater increased the plants height, and grain weight of sorghum.

Key words: Irrigation; Treated wastewater reuse; Chemical soil properties; Sorghum; Gaza

29

1. INTRODUCTION

Climate change and rapid population growth increase water demand in arid and semi-arid

regions like the Mediterranean region, which is considered as area with limited fresh water

resources. In addition, the Mediterranean climate is characterized by hot and dry summers and

mild winters which receive the major part of the annual precipitation. Also, rainfall is unevenly

distributed in space and time (Kamizoulis et al., 2003). Palestinians in West Bank and Gaza

Strip suffer from water scarcity. The dominance of the Israeli occupation over the Palestinian

water and land resources exacerbates demands on limited fresh water supplies (Abu-Madi et

al., 2009).

Therefore, there is an urgent need to conserve and protect fresh water resources and to use the

water of lower quality for irrigation (Rusan et al., 2007). The use of treated wastewater (re-

claimed water) for irrigation could be one of the main options to develop water resources in the

Gaza Strip as it represents an additional marginal and reliable water resources. Using treated

effluent for agricultural purposes as an alternative of groundwater would reduce the deficit and

would reduce the degradation of the ground water quantity (Nassar et al., 2010) by replacing a

sizable volume of fresh water and saving it or diverting it to other uses or sectors in need.

Sorghum [Scientific name: Sorghum Bicolor L. Moench] Sorghum is widely used for food

and as fodder all over the world and it has about the same nutritive quality as corn for domestic

animals. It has the ability to tolerate and survive under adverse conditions of intermittent and

continuing drought. Sorghum has also received considerable attention during the last years as

an alternative source for energy production (Ali Khan et al., 2010). Sorghum crop is considered

a moderate salt tolerant crop (Electrical Conductivity EC = 3-4 dS/m) and has a water require-

ment of about 450–650 mm/growing season, and it`s boron tolerance is (4.0-6.0 mg/l) (Pescod

M.B., 1992).

2. METHODS

2.1 Experimental set up and procedure

The field experiment was conducted from May to September 2011, in Al-Zaitoun agricultural

farm, on an area around 30 m2, about 1.5 km eastern of Gaza Wastewater Treatment Plant

(GWWTP), and located beside Salah Eddeen road, as a pilot study. The climate is Mediterra-

nean Sea climate with warm dry summers and mild rainy winters. The average daily mean

temperature during the study period (May to September 2011) was 28.5°C.

30

The wastewater effluent translocated from GWWTP by pipes to the farm. During the experi-

ment, the effluent was stored in a collection basin with a capacity of 200 CM. The effluent

passed through a sand filter and a Disk filter then to the field. Sorghum Crop was selected for

field experiment.

2.1.1 Experiment design

An experimental site of 26 m2 was prepared to carry out the experiment. The site was divided

into two main plots; each area is 13 m2. The first plot was irrigated with effluent from GWWTP

and the second plot was irrigated with fresh water from agricultural well. Notice that no ferti-

lizers or soil conditioners were used during the experiment period. Chisel ploughing tillage was

used before crop seeding. Drip irrigation system was applied in the field. The sorghum was

sown in May 2011 manually using two seeds per hole; the seed was covered well with soil to

aid germination and the plants were harvested in September 2011.

2.2 Sampling and analysis

2.2.1 Water Sampling

Treated wastewater used for irrigation was collected in polythene bottles from the study area.

Similarly, groundwater samples were also collected from agricultural well used for irrigation

in the study area. The used polyethylene bottles had been pre-washed with acid and distilled

water and then were dried. The samples were preserved at 4˚C in an ice box and brought to the

laboratory at Bir Zeit University Testing Laboratories (BZUTL). The pH and EC were deter-

mined immediately after collection in situ by pH meter and EC meter respectively (Motsara &

Roy, 2008).

Heavy metals like Fe, Mn, and Zn tested by Inductively Coupled Plasma Mass Spectrometry

(ICP) at Heidelberg University Testing Laboratories. Sodium (Na) and Potassium (K) values

were measured by flame photometer, while chloride (Cl) was measured by titration with

AgNO3. Total phosphate (TP) measured by spectrophotometric method.

Biochemical Oxygen Demand (BOD5) , indicates the amount of water dissolved oxygen con-

sumed by microbes incubated in darkness for five days at an ambient temperature of 20°C,

while Total Suspended Solids (TSS) was determined by tests for "total suspended non-filterable

solids", gravimetric method (APHA, 2005).

Total coliforms (TC) and fecal coliforms (FC) were measured by filtration of 100 ml sample

through a 0.45 μm Millipore membrane filters and the filters were incubated for 24 h at 37oC

for TC and 44.5oC for 24 h for FC.

31

2.2.2 Soil Sampling

Composite soil samples (0–30 cm) depth were collected using Auger along systematic paths

(systematic sampling) from plots. Samples were collected before seeding the crop in May 2011

and after harvesting it in September 2011.

Soil samples about 1 Kg were cleaned off stones and plant materials manually and then were

air dried, ground and passed through 2 mm sieves and kept for physical and chemical analysis

at BZUTL. The pH was determined using glass electrode. Soil salinity was determined in soil

extracts using conductivity meter and expressed as EC. Soluble cations (K, Na) were deter-

mined by Flame photometer.

Total alkalinity (CO3-2 and HCO3-), Cations (calcium and magnesium) and Available phos-

phorous were determined using the methods given by APHA (2005). Cl was titrated by standard

silver nitrate and all heavy metals concentrations were determined by ICP after digestion.

2.2.3 Plant Sampling

In each site, the second leaf from the top of 30 plants were randomly selected. Plant samples

analyzed at Greater Cairo Water Company. Leaves were separated and washed with tap water,

rinsed with distilled water and dried at 80˚C for 24 hours. Dry mass of the leaves and stems and

roots was recorded after oven-drying the samples for 72 hours at 80˚C. Dried plant samples

were ground and retained for mineral analysis (Amiri et al., 2008).

3. RESULTS & DISCUSSION

3.1 Characteristics of water used for irrigation and its impact

A pH of 7.6 was found for the treated wastewater, a pH range of 6.5 to 8.4 is desirable for

effluent quality for irrigation according to FAO-1985 guidelines, and a pH range of 6-9 is de-

sirable for effluent quality for irrigation according to the Palestinian Standard for the Treated

Wastewater PS-742-2003. Therefore, the used treated wastewater is suitable for irrigation in

term of pH.

Moreover, Total Dissolved Solids (TDS) was found to be 2300 mg/L, and EC was found to

be 3.83 dS/m; thus, its severe saline, FAO 1985, recommends EC (0-2.0 dS/m) for wastewater

to be used for irrigation. But as shown EC of treated wastewater is above the range, so the

effluent may cause slight to moderate problems. PS-742-2003 recommends TDS of 1500 mg/l

for treated wastewater to be used for irrigation, and excessive salinity may damage some crops.

32

Na concentration for fresh water equals 425 mg/l, and for treated wastewater equals 563 mg/l.

According to PS-742-2003. Both types of water was unsuitable for irrigation due to the high

Na content. The higher concentration of Na in wastewater is due to household products for

laundry, kitchen, bath, and cleaning (Shomar et al., 2005).

Cl concentration of treated wastewater is 760 mg/l, which also makes it suitable only of tol-

erant crops. According to PS-742-2003 the recommends Cl concentration for treated

wastewater to be used for irrigation is 500 mg/l. Treated wastewater has a high concentration

of Cl because it is not removed by wastewater treatment.

Treated wastewater often contains high concentration of nutrients essential to plant growth

such as nitrogen (N), K and phosphorus (P). Significant quantities of a nutrient can be applied

to a crop when irrigated by treated wastewater. Therefore, using treated wastewater will reduce

usage of chemical fertilizers that contribute to various environmental hazards.

Total phosphorous is used as an indicator of pollution from run-off in agricultural areas or

domestic sewage; natural phosphate is due to decayed organic matter and phosphate minerals.

P is also a primary macronutrient that is essential to the growth of plants and other biological

organisms but quantities can be excessive and if the concentrations in water are too high, nox-

ious algal blooms can occur.

Total Kjeldahl Nitrogen (TKN) is an analysis to determine both the organic nitrogen and the

ammonia nitrogen. The analysis involves a preliminary digestion to convert the organic nitro-

gen to ammonia, then distillation of the total ammonia into an acid absorbing solution and de-

termination of the ammonia by an appropriate method.

Results indicated that nitrates (NO3) values in treated wastewater is higher than fresh water

due to the efficiency of treatment plant as the organic load increases with time. Moreover as it

was stated high NO3 has no severe impact on crops but it may leaches to the ground aquifer

(Abu Nada, 2009).

Results shows that treated wastewater suitable for irrigation according to BOD value, and it

is in the range of PS-742-2003. Also, treated wastewater TSS equal 70 mg/l which is above the

PS-742-2003, this is may cause soil plugging in irrigation systems. Results of treated

wastewater shows that FC equal 300 CFU/100 ml. That means lower risk of pathogens pres-

ence, and it is lower than PS-742-2003.

Table (1) shows that Fe, Mn, Zn concentration in treated wastewater is higher than in fresh

water because they cannot be degraded in the WWTP, so they may affect soil and crop, but they

are still under PS-742-2003 values. Some heavy metals are essential to plant growth at low

concentrations. Nevertheless, these heavy metals become toxic and harmful at high concentra-

tions.

33

Table 1. Characteristics of irrigation water used in the experiment.

Fresh wa-ter

Treated Wastewater

PS-742-2003*

pH 6.

8

7.6 6-9

TDS (mg/l) 1

700

2300 1500

EC (dS/m) 2.

83

3.83 -

Na (mg/l) 4

25

563 200

Cl (mg/l) 5

36

760 500

TKN (mg/l) 1

4 60 -

TP (mg/l)

0.

5 3.2 -

K (mg/l) 1

3 31 -

BOD as O2 (mg/l) 1

8

60 60

NO3 (mg/l)

5

5 60 50

TSS (mg/l) 2

7

70 50

TC (CFU/100ml) 6

0

1000 -

FC (CFU/100ml) N

ill

300 1000

Fe (µg/l) 6

5

106 5000

Mn (µg/l) 1

7

68.8 200

Zn (µg/l) <

5

53 2000

* PS-742-2003 for dry fodder irrigation.

34

3.2 Chemical content of the soil at the experimental site

The results show that soil irrigated by both types of water has no difference in pH value and

there is no significant increase after irrigation (7.9 - 8.15). In general soil in the experiment was

moderately alkaline which indicates that free lime (CaCO3) exists in soil according to (Hach

Company, 1992).

The results show that soil after irrigation by fresh water have EC = 1 dS/m and SAR value

decreased to 3.6, but EC of soil after irrigated by treated wastewater slightly decreased to 1.16

dS/m and it is suitable concentrations for sorghum which is tolerant salinity to 10 dS/m as

shown in Table (2). SAR value for soil irrigated by treated wastewater also decreased to 4.3.

So according to Table (2) soil salinity normal in both plots. Organic Matter (O.M%) tends to

increase after irrigation by treated wastewater due to the high content of nutrients result in rich

biomass production, showing a benefit to the soil. Physical and chemical properties are im-

proved due to soil O.M. This enabled granular structures to form, and crop growth accelerated

and enjoyed nutrient absorption (Wang & Wang, 2005).

Results showed that treated wastewater increased significantly the soil Cation Exchange Ca-

pacity (CEC), whereas irrigation with fresh water slightly increased the soil CEC. According

to Donald (1995), CEC less than 3 cmol/kg in sandy soils corresponds with low organic matter,

while CEC of the sandy soil higher than 25 cmol/kg corresponds to high organic matter. In

General soil macronutrients (N, P, K) concentration increased by irrigation with treated

wastewater rather fresh water. Moreover, results shows that micronutrients concentration was

higher in the soil irrigated by treated wastewater, especially, Fe and Cu concentration which

increased after irrigation.

According to the USDA soil textural triangle, the soil texture of the experiment was Sandy

Loam (Sand= 73%, Silt= 20.2%, Clay= 6.8%) and there has been no significant change in the

texture of the soil.

35

Table 2. Soil properties before and after irrigation by treated wastewater and fresh water

Before Irrigation

After Irrigation By

Treated wastewater

Fresh wa-ter

pH 7.97 8.10 8.15

EC (dS/m) 1.1 1.16 1.0

SAR 9.6 4.3 3.6

O.M (%) 0.22 1.0 0.74

CEC (meq/100g)

12.70 13.40 12.97

TN (mg/kg) 196 672 221

TP (mg/kg) 1.4 17 5

K (mg/kg) 3 10 5

CL (mg/kg) 250 180 150

NO3 (mg/kg) 51.5 25 23

Fe (mg/kg) 9.32 11 10.36

Cu (µg/kg) 27.5 73.4 52.05

H.M after digestion.

3.3 Plant analysis

Micronutrients in sorghum crop:

The high concentration of micronutrients in soils was reflected by higher concentrations of

metals in sorghum crop. The results indicated that the highest amount of micronutrients in sor-

ghum irrigated by treated wastewater was (Fe, Pb, Zn, Mn, Cu) according to Table (3).

Table 3. Average concentrations of heavy metals in Sorghum leaves at the end of irrigation period.

Irrigation by treated

wastewater

Irrigation by fresh water

Fe (µg/kg) 5.3 3.5

Pb (µg/kg) 3.5 0.5

Mn (µg/kg) 0.7 0.1

Cu (µg/kg) 0.54 0.45

36

Plant height of sorghum crop:

Results show in Table (4) that the height average of sorghum irrigated by treated wastewater is larger than sorghum irrigated by fresh water due to the quality of treated wastewater enriched

by nutrients.

Table 4. Height average of Sorghum irrigated by treated wastewater and fresh water.

Height (cm) after

month of planting

Height (cm) after two

month of planting

Fresh water 59.37 244.15

Treated

wastewater

64.46 266.21

Total yield of sorghum crop:

Table (5) shows that treated wastewater increased sorghum yield because it provided the soil

with nutrients, which enhanced required for plant growth and soil fertility.

Abu Foul et al., 2014 also have demonstrated similar results from irrigation with treated

wastewater on Plant Height, Thickness, and Number of Leaves of corn plants.

Table 5. Yield of Sorghum crop.

Yield

Kg/plot Kg/ha

Fresh water 6.5 5000

Treated

wastewater

9.5 7307

- Area of plot =13 m2

4. CONCLUSIONS

Treated wastewater has major benefits since it can be an alternative irrigation source to fresh

water resources. In addition, treated wastewater effluent could be suitable for sorghum irriga-

tion which moderate salt tolerant without causing significant heavy metals pollution to soil and

crop. However, Irrigation by Treated wastewater increased the percentage of O.M %, CEC, K,

TP, Ca, Mg, Na, and Cl in soil solution and reduce fertilizer usage. The vegetative growth and

yield of sorghum are enhanced by the application of treated wastewater compared to fresh wa-

ter.

37

References

1. Abu Foul, A., Alkhatib, R., Abu Alqumboz, M., Abunama, T. (2014) Effects of irrigation

with treated wastewater on the physical properties and chemical composition of white corn.

IUG Journal of Natural and Engineering Studies, Vol.22, No.1, pp.77-96.

2. Abu-Madi, M., Mimi, Z., and Sinokrot, N. (2009) Building a participatory national consensus on wastewater reclamation and reuse in Palestine. Second International Conference on the

Palestinian Environment, An-Najah National University, Nablus, Palestine . Abu Nada, Z.

(2009) Long term impact of wastewater irrigation on soil and crop quality parameters in Gaza

Strip. Unpublished Msc. Thesis, Islamic University of Gaza, Palestine.

3. Ali khan, M., Shaukat, S., Hany, O., and Jabeen, S. (2010) Irrigation of sorghum crop with

waste stabilization pond effluent: growth and yield responses, Pakistan, Journal of Botany,

Vol. 42, Issues 3, pp. 1665-1674.

4. Amiri, S., Maralian, H., and Aghabarati, A. (2008) Heavy metal accumulation in Melilotus

officinalis under crown Olea europaea L forest irrigated with wastewater, African Journal

of Biotechnology, Vol. 7, Issues 21, pp. 3912-3916.

5. APHA (2005) Standard Methods for the Examination of Water and Wastewater, 21st ed., American

Public Health Association, Washington, D.C.

6. Donald, S Ross. (1995) "Recommended methods for determining soil cation exchange ca-

pacity." In: Recommended soil testing procedures for the northeastern United States, (J T

Sims & A Wolf), pp.62-69, Newark DE: University of Delaware, Agricultural experiment

station.

7. Hach Company (1992) Soil and irrigation water interpretation manual. Hach, USA.

8. Kamizoulis, G., Bahri, A., Brissaud, F., and Angelakis, A.N. (2003) Wastewater recycling

and reuse practices in Mediterranean region: recommended guidelines.

9. Motsara, M., and Roy, R. (2008) Guide to laboratory establishment for plant nutrient analy-

sis, FAO Fertilizer and Plant Nutrition Bulletins, Rome, Italy.

10. Nassar, A., Tubail, Kh., Moritz, A., and Hall, J. (2010) Attitudes of farmers towards effluent

reuse in the Gaza Strip. Journal of Environmental Sciences, Vol. 39, pp.19.

11. Pescod, M.B. (1992) Wastewater treatment and use in agriculture, FAO irrigation and drainage

paper 47, FAO publication, Rome.

12. Rusan, M., Hinnawi, S., and Rousan, L. (2007) Long term effect of wastewater irrigation

of forage crops on soil and plant quality parameters. Desalination, Vol. 215, pp.143– 152.

13. Shomar, B., Muller, G., Yahya, A. (2005) Geochemical features of topsoils in the Gaza

Strip: Natural occurrence and anthropogenic inputs, Environmental Research, Vol. 98, pp.

372–382.

14. Wang, Q. K., and Wang, S. L. (2005) Forming and stable mechanism of soil aggregate and

influencing factors[J]. Chinese Journal of Soil Science, Vol. 36, Issue 3, pp. 415–421.

Land Reclamation using Construction Waste for Offshore

Gaza Fishery Port

Mazen Abualtayef *2 Hassan Ziara 2 and Ahmed Khaled Seif 3

Received Feb 2018; accepted May 2018

Abstract :The world experience rapidly growing in population density and lack of areas especially coastal areas, which is considered the most vital, economic and cultural areas around

the world. This force the experts investigating different suggestions such as sea reclamation and

its exploitation for many goals.

Gaza Strip is considered one of the most densely area in the world. Where the population is

nearly two million inhabitants by 2016 while the area is 365 square kilometers. This reflects on

the availability of lands in the future that will raise the lands prices. Moreover, the existing fishery harbor constructed in 1994–1998 period has locally disturbed the coastal erosion and

sedimentation pattern and resulting in sand erosion problems. On the other hand, two million

ton of debris have been accumulated in the last aggression on Gaza Strip in 2014. This massive

volume of concrete rubble is considered huge burden on the landfills in the Gaza Strip, which are already overloaded. In this study, an investigation of the best way to dispose these debris by

land reclamation in Gaza Strip.

In this study, it is proposed to relocate the construction features of the Gaza fishery port for better sediment transport and hydraulic conditions. Therefore, the study methodology started

with estimation of the construction waste quantity resulted from 2014 aggression, then ensuring

its testing results, specifications and possibility of using it in land reclamation. Furthermore, the existing Gaza fishery port sediment transportation and features were studied and its bathymetry

was identified. As a result, the new proposed reclaimed area was identified in west of the existing

western breakwater at Gaza fishery port and its area was estimated according to estimation the

quantity of debris resulted from removing the existing breakwaters in addition to construction wastes resulted from 2014 aggression. The total estimated quantity of available debris is about

one million cubic meters, which adequate to reclaim about 11 hectares. However, the new

reclaimed area should be surrounded by sheet piles, so the sheet pile type is assumed to be PZC 28 and finally the total cost of the proposed reclaimed area was estimated as USD130 per square

meter of reclaimed area. Finally, to facilitate transportation and movement through this

reclaimed area, the bridge is recommended to be constructed. Finally, the cost of sea

reclamation is feasible especially that Gaza fishery seaport is very vital and important place, which will be as a fishery port and for recreational activities.

Keywords: Land reclamation, construction waste, Gaza fishery port

2 Environmental Engineering Department, the Islamic University of Gaza, Palestine) mabualtayef@iugaza.edu.ps 2 Municipality of Gaza, Palestine 3 Department of Civil Engineering, Al-Azhar University, Egypt

39

1. INTRODUCTION

More than one-third of the world’s population resides in coastal areas, which account for just

4% of Earth’s total land area. Coastal population densities are nearly three times that of in land

areas and are increasing exponentially. Coastal human settlements usually exploit their position by

reclaiming tidal and shallow sea areas through land reclamation. This phenomenon can be observed

in many coastal countries and cities, such as Korea, Japan, Singapore, the Netherlands, Hong Kong

and Macau (Feng et al., 2014).

Reclamation in coastal zones is effective for relieving population pressure and ensuring food

safety. Since the 1950s, the development of coastal zones has entered a peak period. At present, the

reclamation of coastal zones mainly occurs in developing countries. The coastal reclaimed lands

are mainly used for agricultural production, urban and industrial development, and port construc-

tion (Li et al., 2014). Land reclamation is a process to create new land from the sea, which can be

achieved with several different methods. The simplest method involves simply filling the area with

large amounts of heavy rock and/or cement, then filling with clay and dirt until the desired height

is reached (Nadzir et al., 2014).

As coastal area is a very sensitive area, any development needs to be highly evaluated for its

possible disturbances. It is because the coastal reclamation comes with its adverse impacts to the

land. Hazard in the coastal area found through erosion activity and caused by environmental change

and human actions. If ecosystem undermined, the ability of the coastal areas to adapt and regenerate

would erode (Nadzir et al., 2014). The conversion of sea to land permanently changes the natural

characteristics of the ocean and coastal environment and cause considerable damage to the marine

ecosystems upon which human kind depends. The impacts of reclamation not only limited to the

area where damped/dredged and reclaimed occurred, but impacts felt over a larger area where sil-

tation or change in current happened (Azwar et al., 2013).

This study investigates in focus the feasibility of sea reclamation practices in the Gaza strip where

due to the lack of lands for several activities pushes the decision makers toward sea reclamation.

Gaza strip is considered as one of the highly populated density areas around the world, where the

population is nearly 2 million inhabitants by 2016 while the area is 365 km2 (PCBS, 2016). The

implementation of such reclamation projects can be considered as an urgent to Gaza strip, for a

number of reasons. The main of these reasons are the increasing in population growth, the economic

recession and lack of areas.

Indeed, the coast of Gaza was affected by man-made structures such as the construction of fishery

harbor, which started in 1994 and completed in 1998. The fishery harbor has locally disturbed the

coastal erosion and sedimentation pattern and resulting in sand erosion problems. Furthermore, the

building and roads adjacent to the shoreline are facing a stability problem and it is expected to have

40

a serious erosion problem in the coming few years specially in the region of Beach camp that locates

to the north of the port`s site (Abualtayef et al., 2012). As a counter-measure to this, the construction

waste was deployed in the eroded area, which works as a beach revetment, to mitigate the severe

beach erosion and protecting the hotels. UNRWA has constructed gabions along the Beach camp

with a total length of 1650 m to protect 3 the main coastal road. Several short groins have been

constructed along the Beach camp for shoreline preservation. Actually, these mitigations are not

effective and hence significance measures should be under taken to protect the beaches against

coastal processes due to the fishery harbor (Abualtayef et al., 2012).

On the other hand, due to the three aggressive invasions that occur in Gaza strip in 2008, 2012,

and 2014, about two million tons of debris have been accumulated in the lands from damaged

buildings and facilities. However, the last war on the Gaza strip was one of various rationales that

played big role in deteriorating infrastructure conditions in the strip. The ten-year closure had al-

ready left most of infrastructure facilities inadequate to function. In particular, municipal services,

especially solid waste and solid waste treatment, had to be curtailed, leading to the accumulation

of hundreds of tons of rubbish on the streets each day. Restrictions on the imports of essential

consumables and other materials also reduced the efficiency of the operation of sanitary landfills

and garbage collection trucks. (UNDP, 2014).

To overcome solid waste storage problem, Municipality of Rafah was the first who carried out

crushing activities in relatively large quantities. Funded by Italian Government, the municipality

was supplied by a small-scale crusher capacity of 70 tons per hour and started crushing of concrete

rubble generated in the south of the Gaza strip. The produced crushed material was used by the

municipality in agricultural roads. Later, a small quantity was used by UNRWA in some roads in

Tal Sultan area in Rafah. The next large-scale crushing of concrete rubble was followed by UNDP

after disengagement of Israeli occupation from Gaza settlements. In 2006, UNDP was assigned by

quartet to remove and crush more than 700,000 tons of mixed concrete rubble from Gaza settle-

ments. Nearly 400,000 tons of this rubble was removed in very good and clean conditions (UNDP,

2014).

In general, two million tons of construction waste that was generated from the last war on Gaza.

One of these problems is how and where to dispose this massive volume of concrete rubble taking

into consideration that almost all available landfills in the Gaza strip are already overloaded. The

Palestinians ministries proposed many ideas to effective disposal for this debris. In this study, an

investigation of the best way to dispose these debris by land reclamation in Gaza strip especially

that the huge need to mitigate the problem of the existing Gaza fishery port.

41

2. MATERIALS AND METHODS

Producing of reusable crushed material was an essential step to obtain the best benefits from gen-

erated post-war rubble such as reducing the total volume of the rubble in already overloaded land-

fills and bridging the gap between demand and supply of construction aggregates in construction

demand industry taking into consideration performing required tests that approve the application

of such recycled materials.

In 2009, the quantities of debris were gathered from UNDP and other relevant authorities in the

Gaza strip. Then, the samples were taken by two teams from IUL and AEL labs arranged for taking

a sample from 30,000 tons of crushed materials from post-war rubble. The objective of testing

crushed materials was to determine the technical applicability of using the recycled concrete rubble

collected from post-war affected sites in Gaza strip in road construction as an alternative for the

natural 4 aggregate in road construction or other applications. The characteristics of such aggregates

were determined and compared to international standards. The reuse alternative was investigated

in road and concrete construction throughout all performed tests. The test analysis showed that the

recycling of the concrete rubble aggregates and its use in road sub-base give acceptable results.

Thus, recycled aggregates can be considered as a good alternative to natural aggregates especially

in road constructions.

2.1 Site bathymetry

The bathymetric features of the Gaza fishery port were gathered by real field survey using

sonar as it shown in Figure 1.

Figure 1: The bathymetric features of the Gaza fishery port

42

2.2 Materials and rubble quantities

The 51-day, July-August 2014, military operation in the Gaza strip has caused the destruction of

infrastructure and buildings. In order to verify the preliminary infrastructure damage assessment

findings and to further inform on the actual damages, the Higher Inter-Ministerial Committee

tasked UNDP to conduct a detailed infrastructure damage assessment in collaboration with line

ministries, UNRWA, UNOSAT and WFP. UNDP estimated that around two million tons of rubble

have been generated, which is three times more than the amount of rubble generated during 2008-

09 Gaza war. The detailed quantity of generated rubble according to the Gaza strip governorates is

shown in Table1 (UNDP, 2016).

The analysis results of specific gravity for the crushed material is 2.35, accordingly the

available volume of filling material for reclamation is about 850,000 m3 (UNDP, 2016).

2.3 Characteristics of debris

UNIDO (2005) conducted a testing program to investigate the application of construction and

demolishing wastes in construction industry in the Gaza Strip. The performed testing program

aimed to highlight the possibility of producing recycled aggregates from the construction and dem-

olition wastes (CDW) and was performed on a sample taken from concrete rubble in Rafah area.

The reuse alternative is investigated in concrete mixes and road construction throughout compre-

hensive testing program. The test results showed that the recycling of the CDW aggregates and its

use in both concrete and road subbase gives acceptable results.

Furthermore, UNDP (2009) conducted testing program on samples were taken from concrete rub-

ble that collected from post-war rubble. The results show good opportunities for using crushed

concrete rubble in construction industry. Many tests were performed taking into consideration pre-

vious international experience in this field where more than 900 million tons of concrete rubble is

annually generated and partially reused in USA, Europe and Japan.

Sieve analysis: The collected samples of crushed concrete rubble were sieved, and the results

were compared with the standard limits of AASHTO for base course and sub-base materials grade

43

(A). as a result, the samples showed that they were going down to lower standard limit which rep-

resents the course limit. Some of the samples were courser than the standard limits and others were

slightly matching these limits. From technical point of view, this gradation is acceptable to some

extent. The large particles, greater than 2.5 cm are suitable for road applications However, for con-

crete application it is recommended to use small particles, smaller than 2.5 cm. For concrete appli-

cation it was recommended to conduct three tests: compressive strength test at 7 and 28 days, slump

test and air content test. Physical properties of these fractions as obtained from previous studies are

shown in Table 2 (El Kharouby, 2011).

Analysis of gradation: Results of the sieve analysis road applications showed that the

crushed material is classified as coarse material greater than 4.75 mm (sieve no. 4). As shown

in Table 3, the coarse to fine materials ratio was on average 77% to 23%, respectively (El

Kharouby, 2011). The amount of course materials according to AASHTO should not exceed

70% and for fine materials 40%. This means that an additional amount of fine materials

should be added to meet the standards. Table 4 6

shows the test results for other essential requirements of crushed concrete comparing to in-

ternational standards (El Kharouby, 2011).

Recycled concrete rubble seems to have satisfying properties for the most common exposure

conditions. It can solve many of the basic problems concerning shortage of construction

materials in roads and concrete construction and reclamation. In addition, as natural resources

diminish, the demand for recycled concrete aggregate is likely toincrease, making concrete

44

recycling the economically and environmentally preferable alternative to traditional “smash and

trash” demolition. Wherever good natural aggregates are not locally available, where natural

aggregate costs exceed removal and recycling costs or where disposal of existing concrete

pavement or concrete structures is problematic, concrete recycling should be evaluated. Moreover,

concrete recycling appears to be profitable. In most cases, it can meet demand requirements of

lower value product applications such as land reclamation.

2.4 The study methodology

The characteristics results of concrete rubble showed the possibility of using it in sea recla-

mation. Therefore, sea reclamation will be the best solution to alleviate these problems. The

study methodology was implemented as in the following steps: estimating the total quantity of

rubble resulted from the 2014 war on Gaza strip, testing characteristics of concrete rubble for

sea reclamation possibility, studying the existing fishery harbor area and characteristics to de-

termine the suitable interventions and which tongues will be removed, defining the bathymetric

features of the Gaza fishery port by real field survey using sonar to determine the sea water

depth at different points as it shown in Figure 1, estimating the rubble quantity resulted from

removing the existing breakwaters of the Gaza fishery harbor, estimating the total rubble quan-

tity will be reclaimed by adding the previously estimated removed rubble quantity to the re-

sulted from the 2014 war on Gaza strip. Defining the proposed reclaimed area dimensions and

estimating its area as it shown in Figure 2, choosing the suitable sheet pile needed to reclama-

tion based on the soil type, seawater depth and loads; cost estimation of the proposed area

reclamation, and finally defining the trucks movement and routes during reclamation imple-

mentation.

45

46

Figure 2: The proposed reclaimed area for offshore Gaza fishery port

3 RESULTS AND DISCUSSION

Gaza fishery harbor considers the only port for the ships and boats of Gaza’s fishers, but unfortu-

nately this port causes significant problems of erosion/accretion along Gaza’s coastline. In this

regard, the coastal researchers studied in deep several alternatives to relocate/redesign the port in

order to mitigate its impacts on the coastlines and nearby structures. 8

Abualtayef et al. (2013) recommended that relocating the current situation of Gaza fishery harbor

into offshore fishery harbor is the most suitable alternative to mitigate the erosion/accretion im-

pacts, but the main obstacle of this design is the generation of strong current of 1.0 m/s at the

entrance of the harbor. However, this obstacle can be overcome by some arrangements of struc-

tures. So, this section explains and discuss the estimation process of the proposed reclaimed quan-

tity and area in addition to estimate the reclamation cost.

3.1 Existing breakwaters quantity estimation

The quantity of construction wastes that intent to be used in relocating the current design into

offshore fishery harbor is 850,000 m3. Fortunately, the reclaimed area can be increased if we ex-

ploit the used materials in the 300 m and 500 m breakwaters as shown in Figure 3.

To calculate the volume of rubble resulted from removing the 300 m and 500 m breakwaters

extended from the beach, it assumed that this area is divided to three different segments areas;

A, B and C, then their dimensions is shown in Figure 4. After that, the area of each segment is

calculated as shown in Table 5.

47

Figure 3: The existing Gaza fishery harbor

(source: Google earth, 2016)

Based on the calculated volume for each segment, the total volume of rubble produced from re-

moved these three segments is 204,800 m3. In addition, the available volume of rubble accumu-

lated from the last 2014 war on Gaza Strip is approximately 850,000 m3. Therefore, the total

available rubble volume for proposed reclaimed area is more than 1.0 Million m3.

48

3.2 Estimation of the proposed reclaimed area (new offshore Gaza fishery port)

The required area for proposed reclamation is calculated by using the total amount of produced

rubble, which previously estimated as 1,053,000 m3.

According to the contour map is shown in Figure 1), the seawater depth at 500m, 600m and

700m away from the beach are -7m, -8m and -9m respectively. However, the existing breakwater

ground is 2m above sea water level.

To calculate the required area as it shown in Figure 5, the following equation is used:

The volume of new reclaimed area =

d1 is the seawater depth at 500 m away from the beach, which is equal to 7 m

d2 is the seawater depth at the end limit of the proposed reclaimed area

X is the width of the reclaimed area

L is the breakwater length (also the length of the reclaimed area), which is 500 m

by substitution with these values and using the volume of new reclaimed area of 1.0 million m3,

the resulted equation with two variables is solved by goal seek on excel program and keeping the

minimum width of the reclaimed area 200 m. the resulted value of d2 and x are 9 m and 208 m,

respectively.

49

As the width of existing breakwater segment D (Figure 4) is 20m, the total width will become

228 m and length of 500 m.

3.3 Cost estimation of proposed reclaimed area

Rubble transportation cost estimation

Based on the coordination with rubble relevant sectors, the accumulated rubble from the last

war in 2014 on Gaza Strip is considered available. So, the accumulated rubble will be trans-

ported to the project site by assumption that the truck capacity is 16 m3. As the cost of one

truck trip is about 65$ and as the total quantity need to transport is 1,000,000 m3, so the re-

quired trips number is 65813 trips. The total transport cost is 4.3M USD. The filling and

damping cost is estimated to be 200% of transport cost, so the total needed cost 8.6M USD.

Sheet piles cost

It is assumed that the need sheet pile section is PZC 28 as it shown in Figure 6.The weight

of 1 square meter of PZC 28 section is 166.1 kg/m2. The reclaimed area shape is rectangular,

with the total perimeter of 1,456 m ([500+228]×2). Assuming the average seawater depth at

the new reclaimed area is 8 m and as the existing breakwater ground is 2 m above seawater

level, the sheet pile will be installed at 5 m underground, so the total height of the installed

sheet pile is 15 m. The total weight required of sheet pile 3630 tons (1457×15×166.1). The

cost of PZC 28 sheet pile section is about 550$/ton. So, the total cost of need sheet pile steel

is 2M USD.

50

Sheet pile installation cost

In common, the sheet pile installation cost at the site is about 300% of the sheet pile steal

cost. So, the installation cost is 6M USD.

Finally, the total cost of the new reclamation area is equal to the summation of rubble

transportation to the site project cost and installation of sheet pile steel cost as shown in Table

3. Therefore, the total project cost is 14.6M USD. In conclusion, as the total reclaimed area is

114,000 m2 (500×228), the cost per square meter of reclaimed area is 130 USD/m2.

51

3.4 Proposed reclamation process

Defining the proposed reclaimed area that is the area located in the west of the existing

western breakwater.

Figure 7: Illustration of the proposed reclaimed area location and method

(source: DEME, 2014)

Choosing the sheet pile needed to reclamation based on the soil type, seawater depth and

loads. Therefore, the chosen sheet pile type is hot rolled steel sheet pile PZC 28. The sheet pile

installation will be around the entire proposed reclaimed area.

52

Figure 8: Installation of sheet piles

(source: WIKI, 2016)

Defining the trucks movement and routes during reclamation implementation, which will

be through segment B to segment C to fill the area located on front of segment D.

Figure 9: The trucks movement during reclamation process

Source: (Yin Pumin, 2014)

The reclamation steps will start with using the rubble resulted from 2014 war on Gaza

Strip, then continuing reclamation by using the rubble from removing the existing breakwater A,

B and C, respectively. Trucks movement during reclamation process is considered, so the remov-

ing will be started by segment A then B and will be finished by segment C.

53

Figure 10: Reclamation activities

(source: Callaert and Van Den Bogaert, 2016)

Construction of bridge to connect the new offshore fishery port to the land. The final

shape of new fishery port will be as in Figure 11 with a rectangular terminal.

Figure 11: The general view of proposed reclamation and bridge installation

(source: Xinhua, 2016)

54

4. CONCLUDING REMARKS

The implementation of such reclamation projects can be considered as an urgent to

Gaza Strip because of the increasing in population growth, the economic recession and lack

of areas.

Land reclamation in Gaza Strip is significant choice to mitigate problems of massive

volume of concrete rubble that was generated from the last war on Gaza especially that almost

all available landfills in the Gaza Strip are already. overloaded. On the other hand, the shortage

of natural aggregate beside the high prices made the recycle of concrete rubble as one of top

priority for land reclamation process.

Sea reclamation is the best solution for many problems, especially for solving the ero-

sion problem at areas north of existing Gaza fishery seaport. Reclamation process will permit

to sediments transport to the northern direction.

The total cost of reclaimed area of 11.4 ha is 15M USD, and the cost of one square

meter of is 130 USD. Based on this result, the cost of sea reclamation is very feasible espe-

cially that Gaza fishery seaport is very vital and important place for fishery and recreation.

55

References

1. Abualtayef, M., Abu Foul, A., Ghabayen, S., Abd Rabou, A., Seif, A., Matar, O. (2013).

Mitigation measures for Gaza coastal erosion. Journal of Coastal Development, 16(2), 135-

146.

2. Abualtayef, M., Ghabayen, S., Abu Foul, A., Seif, A., Kuroiwa, M., Matsubara, Y., Matar,

O. (2012). The impact of Gaza fishery harbor on the Mediterranean coast of Gaza. Journal

of Coastal Development, 16(1), 1-10.

3. Azwar, S. A., Suganda, E., Tjiptoherijanto, P., Rahmayanti, H. (2013). Model of Sustaina-

ble Urban Infrastructure at Coastal Reclamation of North Jakarta. Procedia Environmental

Sciences 17, 452 – 461.

4. Bart Callaert and Jan Van Den Bogaert. (2016). The Taparura Project: Sustainable Coastal

Remediation And Development At Sfax, Tunisia.

5. DEME (2014). Land Reclamation, retrieved from www.deme-group.com

6. El Kharouby, A. (2011). Post-war rubble removal and potential use of recycled construction

rubble in Gaza governorates. The Islamic University Journal, 19(1), 179-212.

7. Feng, L., Zhu, X., and Sun, X. (2014). “Assessing coastal reclamation suitability based on a

fuzzy-AHP comprehensive evaluation framework: A case study of Lianyungang, China”

Marine Pollution Bulletin 89, 102–111.

8. Gerdau (2016). Z-Profile Steel Sheet Piling retrieved from sheet-piling.com

9. Nadzir, N. M., Ibrahim, M., and Mansor, M. (2014). “Impacts of Coastal Reclamation to the

Quality of Life: Tanjung Tokong community, Penang” Procedia - Social and Behavioral

Sciences 153,159 – 168.

10. PCBS (2016). Retrieved from www.pcbs.gov.ps

11. United Nation Development Programme (UNDP): Examining Potential Use of Recycled

Construction Wastes: Analysis Report, Gaza 2007

12. United Nation Industrial Development Programme (UNIDO): Testing Program to Investi-

gate the Application of Construction and Demolition Wastes in Construction Industry in

Gaza Strip: Analysis Report, Gaza 2005.

13. WIKI (2016). Sheet piles retrieved from www.designingbuildings.co.uk/wiki

14. Xinhua (2016). Bridge connects China's first artificial offshore island with mainland re-

trieved from www.globaltimes.cn

15. Yin Pumin (2014). A Sea Besieged, Land reclamation poses great challenges to the coun-

try's ecosystems retrieved from www.bjreview.com.cn

56

Prediction of Hourly Indoor Carbon Monoxide Concentrations by Using Multivariate Methods with Sensitivity Analysis Tech-

nique

Maher Elbayoumi*1 Suheir Elbayoumi Harb2

Received Feb 2018; accepted May 2018

Abstract: Precise site prediction of indoor hourly carbon monoxide (CO) concentrations in

school buildings is a key issue in air quality research nowadays due to its impact on children’s

health. In this study multivariate statistical methods, multiple linear regression (MLR) and

principle component analysis (PCA), were employed to predict hourly indoor CO concentration

in Gaza Strip, Palestine. Measurements were carried in 12 schools from October 2012 to May

2013 (one academic year). The results suggested that the selected models are effective

forecasting tools and hence can be applicable for short-term forecasting of indoor CO level. The

predicted indoor CO concentration values agree strongly well with the measured data with

high coefficients of determination (R2) 0.869, 0.870 for MLR and PCA-MLR (PCR) respectively.

Overall, results showed that PCA model combined with MLR improved MLR model of predicting

indoor CO concentration, with reduced errors by as much as 7.14%.

Keywords: Natural Ventilation; Indoor Air Quality; Data Driven Models; Air Quality Prediction.

1-Maher Elbayoumi, Energy and Sustainable Environment Center, School of Engineering, Israa University, Gaza, Palestine. elbayoumimaher@gmail.com

2-Suheir Elbayoumi Harb, Palestine Technical College, Dier Elbalah, Gaza Strip, Palestine,suheirma@gmail.com

57

1. INTRODUCTION

CO is one of the most characteristic traffic pollutants in urban areas and produce as a primary

pollutant during the incomplete combustion of fossil fuels and biomass in fumes produced by port-

able generators, stoves, and gas ranges (USEPA, 2013). Carbon Monoxide (CO) is colorless and

odorless pollutant that arises from both natural and anthropogenic sources. CO exhibits toxicity

characteristics due to its higher affinity with hemoglobin and as a result reduces the delivering

oxygen to the body’s tissues. Sustained exposure to CO has long been associated with effects on

increasing adverse cardiovascular outcomes, asthma symptoms, hospital admission rates, and heart

rate among children. (Slaughter et al., 2003; Liao et al., 2004; Cakmak et al., 2006; ATSDR, 2012).

There are also adverse impacts for exposure to low concentrations of CO for a long period among

children on learning ability, manual dexterity, attention level, headaches, dizziness, nausea (feeling

sick) and tiredness (HPA, 2009; USEPA, 2013) (Raub and Benignus, 2002; Goniewicz et al., 2009;

HPA, 2009).

Several studies have investigated diurnal and seasonal CO concentration in different type’s build-

ings. However; these studies have mainly focused on the monitoring, but not on the prediction of

IAQ inside buildings (Chaloulakou et al., 2003; Currie et al., 2009; Elbayoumi et al., 2014a). In

addition to that, indoor CO concentrations heavily depend outdoor CO concentrations, indoor

sources, infiltration, ventilation, and air mixing between and within rooms and on local conditions,

such as weather changes,such as differences in temperature, humidity, pressure, atmospheric sta-

bility, and wind speed.Therefore; direct and long-term measurements of CO concentrations are not

practical. In the absence of effective and efficient means to directly measure indoor CO from school

buildings, development of mathematical prediction models might be a good alternative to provide

reasonably accurate estimates.

Multiple linear regression (MLR) is one of the most popular methodology to express response of

a dependent variable of several independent variables (predictor). Several studies used MLR to

correlate the outdoor CO, PM, ozone, NO2 and meteorological variables with indoor concentration

of such pollutant (Chaloulakou et al., 2001; Adar et al., 2008; Braniš and Šafránek, 2011; Elbay-

oumi et al., 2014b). In spite of its success, MLR presents problems in identifying the most im-

portant contributors when multicollinearity, or high correlation between the independent variables

in regression equation are present (Abdul-Wahab et al., 2005). The most common methods for

removing such multicollinearity are principal component analysis (PCA) which has been proven to

be effective tools to study the relationship between voluminous data such as air pollution and me-

teorological records (Yeniay and Goktas, 2002; Poupard et al., 2005; Ul-Saufie et al., 2010). PCA

58

is used to reduce the number of predictive variables and transform into new variables that are mu-

tually orthogonal, or uncorrelated, as well as to determine dominant multivariate relationships (Ab-

dul-Wahab et al., 2005). However, one of the main drawback is that MLR and principal component

regression (PCR) cannot adequately model the non-linear relationships (Al-Alawi et al., 2008).

In the literature, little attention has been paid to forecasting indoor air quality within buildings.

Thus, the overarching goal of this project is to present the results of the application of multivariate

regression analysis (MLR and PCR) in predicting indoor CO concentration as the function of me-

teorological parameters and other pollution concentration from natural ventilated school buildings.

2. METHODOLOGY

2.1 Study Area

The Gaza Strip (365 km2) is located on the eastern coast of the Mediterranean Sea between

longitudes 34° 15' and 35° 40' east, and latitude 29° 30' and 23° 15' north. Climatically, the average

daily temperature fluctuates from 24°C in summer to 15 °C in winter. Meanwhile, the daily relative

humidity varies between 62.5% in the daytime and 83.4% at night in the summer, and between

51.6% in the daytime and 81.3% at night in winter, and the monthly average wind speed is 3 m/s.

(Koçak et al. 2010, PMD 2012). The major source of CO in the Gaza Strip is the exhaust of about

61,000 motor vehicles as for 2012, most of which are more than 15 years old and are out-dated

(PCBS 2012). Exhausts contain large quantities of CO, CO2, PM2.5, and hydrocarbons.

In addition, during the frequent power outages, many people and institutions use portable

electrical generators. Most of the generators used in the Gaza Strip were placed outside but were

very close to the buildings to allow the generators to connect to the central electric panel

(Elbayoumi et al. 2014). CO from these sources can build up in enclosed or partially enclosed

spaces.

Description of Sampling Locations

The concentrations of pollutants were monitored at 12 schools located in north, middle, and south

of the Gaza Strip from October 2012 to May 2013 for one academic year. The sampling schools

were purposely selected to reflect the diverse natures of human and vehicular activities. The

sampling locations are shown in Figure 1. In each selected school, three representative classrooms

were selected for three sampling days. Sampling was conducted both inside and outside the selected

classrooms during the studying activities.

59

Figure 1: Map of the sampling locations (monitoring schools) in the Gaza Strip (Left) as part of

Historical Palestine (Right).

2.2 Measurements and Instrumentation

The measurements were taken place in each site during school hours for three consecutive days.

The samplers were placed inside the classrooms opposite the blackboards at least 1 m from the

wall and at least 1.5 m height from the floor, as recommended by Blondeau (2005) and WHO

(2011). For outdoor sampling, the samplers were placed at the front side of the building, usually

near the playground area. A Kanomax IAQ Monitor was used for measuring CO and carbon di-

oxide (CO2) concentrations. Meanwhile, the ventilation rate (VR) was calculated using the in-

door concentration of CO2 as a surrogate of the ventilation levels per occupant (Kulshreshtha and

Khare 2011, WHO 2011). The mass concentration of particles matters (PM2.5 and PM10) has been

monitored using handheld optical particle counter (HAL-HPC300). The monitor performs partic-

ulate size measurements by using laser light scattering. Air with multiple particle sizes passes

through a flat laser beam produced by an ultra-low maintenance laser diode. A 3–channel pulse

height analyzer for size classification detects the scattering signals.

Meteorological Data

The surface wind speed (WS), ambient temperature (Temp), relative humidity (RH) and dew

point temperature (TDP) in each site were simultaneously measured at the same time with pollu-

tants measurement. A Kanomax IAQ Monitor was used for the temperature and relative humidity

measurements. Smart Sensor Electronic Anemometer was used for wind speed.

Data Interpretation

A vital step in the development of a forecast indoor air model is the choice of input parameters

(Jef et al. 2005). Sensitivity analysis is a very useful method for ranking the importance of input

60

variables by assessing their contribution (percentage) to the variability of the model output. In order

to choose the most appropriate set of inputs parameters for FFBP, a number of statistical methods

can be applied such as stepwise regression (SR), principal component analysis (PCA), and cluster

analysis (Wilks 2011). The importance of these methods is to reduce the number of input variables

into the models, and, thus, considerably diminish redundant information, instabilities and over-

fitting. Here, the selection of variables for the model was made independently for each monitoring

school, through a forward stepwise regression (FSR). During this procedure, which starts with the

variable most correlated with the target, additional variables are added which, together with the

previously selected variables, most accurately predict the target (Wilks 2011). The procedure stops

when any new variable does not significantly reduce the prediction error.

Significance is measured by a partial F-test applied at 5% and by using the standardized

regression determination coefficient (R2) values (Wilks 2011). All 15 potential predictors for indoor

CO were first considered. The use of the FSR for each monitoring school has reduced the

complexity by retaining substantially less variables. The analysis of the data was carried out using

the statistical software, SPSS (Statistical Package for Social Science, version 22) and MATLAB,

version10. The data had been classified randomly into two sets using MATLAB software. Data set

1, which consists of 70% of the original data, was used for model formulation; and data set 2 (30%)

was used for model validation.

Multiple Linear Regression (MLR)

Stepwise multiple regression was carried for CO and the result was checked for multicollinearity

by examining the variance inflation factors (VIF) of the predictor variables. Durbin Watson statistic

used to check if the model does not have any first order autocorrelation problem. MLR can be

expressed according to the following Equation (1):

y = 𝑏0 + 𝑏1𝑥1𝑖 + 𝑏2𝑥2𝑖 + ⋯ + 𝑏𝑘𝑥𝑘𝑖 + ɛ (1)

where, 𝑏𝑘 is the regression coefficients, 𝑥𝑘 𝑖s the explanatory variables, i =1,2,……k and ε is

stochastic error associated with the regression (Agirre-Basurko et al., 2006). The residuals (or error)

were checked to evaluate if they were normally distributed with zero mean and constant variance

to verify the adequacy of the statistical model (Al-Alawi et al., 2008).

Principle Component Analysis (PCA)

PCA is a multivariate technique that is widely used in dealing with a large amount of data in

monitoring studies, such as air pollution studies. In this study, this technique is applied for variables

61

reduction and to provide the most relevant variables in CO variations (Dominick et al., 2012). The

PCs were extracted so that the first principal component (PC1) accounted for the largest amount of

total variation in the data set, whereas the following components accounted for the remaining var-

iations that were not considered in PC1 (Kovač-Andrić et al., 2009). In general, the PCs are ex-

pressed in Equation 2 as follows:

𝑃𝐶𝑖 = 𝑎1𝑖 𝑣1 + 𝑎2𝑖 𝑣2 + ⋯ + 𝑎𝑛𝑖 𝑣𝑛 (2)

where PCi is the principal component i and ani is the loading (correlation coefficient) of the orig-

inal variable V1 (Özbay et al., 2011).

Each PC represents a linear combination of data (variables) at specific coordinates at different

values of chosen parameters (Elbayoumi et al., 2014b). PCs are computed by calculating eigenval-

ues and eigenvectors. Eigenvalues will determine the eigenvectors and PCs; for each PC, only ei-

genvalues larger or equal to 1 are considered significant (Ul-Saufie et al., 2013). Rotated PCs using

varimax rotation to maximize the relationship between the PCs and original variables (Abdul-

Wahab et al., 2005). Dominick et al. (2012) reported that varimax rotation ensures that each vari-

able maximally correlated with only one component and has minimal association with other com-

ponents. The significant variables for each component are determined based on the loading factor

where greater than 0.5 is considered strong, 0.4 is moderate, and 0.30 is weak.

Hybrid Models

Hybrid models are models combine MLR technique with PCA. PCR is a combination of MLR

and PCA. The use of PCs as input in MLR is intended to reduce the complexity and multicolline-

arity problems of the models. The selected variables with high loading from PCA ensured that the

majority of the original variances were included in the models, and they were ideal for use as inde-

pendent variables in MLR (Gervasi, 2008; Gvozdić et al., 2011). Figure 2 shows the architecture

of a PCR model for prediction of indoor CO concentrations.

62

Figure 2. Architecture of a PC-MLR model for the prediction of indoor CO concentrations.

2.3 Performance Indicators

The analysis of prediction performance typically involves calculation of errors between observed

y and predicted 𝑦 values. In this study five performance indicators were used which are normalized

absolute error (NAE), root mean square error (RMSE), predication accuracy (PA), coefficient of

determination (R2) and index of agreement (IA). Normalized absolute error (NAE) and root mean

square error (RMSE) were used to find the error of the model where value closer to 0 indicated a

better model. Meanwhile, the other three performance indicators, i.e. index of agreement (IA),

prediction accuracy (PA) and coefficient of determination (R2) were used to check the accuracy of

the model result, where a higher accuracy is given by value closer to 1 (Gervasi 2008, Karppinen

et al. 2000).

3. RESULTS AND DISCUSSION

3.1 Descriptive Statistics

Figure 3 shows the box plot and descriptive statistics of daily indoor and outdoor CO

concentrations from 2012 to 2013. In December 2012, February and April 2013, both indoor and

outdoor CO concentrations were higher than the remaining months. The outdoor concentrations

were 4.80 ppm, 6.07 ppm, and 5.35 ppm for December, February and April, respectively. CO is

considered an urban-scale pollutant that is generated by road traffic and tends to be present at high

concentrations throughout the city and at significantly reduced concentrations in adjacent rural

areas (WHO 2005). The school locations displayed in Figure 1 are very close to street intersections,

and most of the schools located in overpopulated areas are characterized by congested traffic. Thus,

frequent traffic jams resulting from poorly maintained roads, high traffic density, and very low

63

wind speed are considered the main factors that contribute to high emission, accumulation, and low

dilution of generated CO. Moreover, during the December and February months (December-

January-Febraury are the coldest months of winter in Palestine), the catalytic converters of vehicles

take time to reach the operating temperature when the engine is cold, thereby resulting in increased

CO production (Marković et al. 2008). In addition, the CO production in overcrowded residential

areas increases when cars move slowly near schools. Thus, the location of the investiged schools

may influence indoor CO concentrations. Most of time, the maximum daily concentrations of CO

were below WHO’s guidelines of 9 ppm, except for some exceedences that were observed during

November, December, February, March and April. The total number of exceedences that were

recorded on outdoor concentrations is 82 exceedences, and 22 exceedences on indoor

concentrations. The indoor /outdoor (I/O) ratio were less than 1.0 during the monitoring period.

Similar results were obtained by Chaloulakou and MaBVRoidis (2002), who revealed that air

pollutants, such as CO, that are non-reactive and cannot be absorbed strongly on walls have an I/O

ratio close to 1.0 in the absence of indoor sources. Thus, the buildings’ envelope provides little

protection from outdoor CO pollution, and peaks in indoor concentrations reached the extremes of

outdoor concentrations, regardless of the airtightness in these buildings. Airtightness is the

fundamental building’s property that impacts infiltration and exfiltration. In other words, it is the

64

uncontrolled inward and outward leakage of outdoor air through cracks, interstices or other

unintentional openings of a building, caused by pressure effects of the wind and/or stack effect.

Figure 3: Box plots and descriptive statistics for monthly CO indoor and outdoor for the investigated

schools in the Gaza Strip, Palesti

3.2 Bivariate Correlation Analysis

Several studies confirmed that the IAQ is dependent on outdoor concentrations and local condi-

tions, such as weather changes and seasonal variations (Kam et al. 2011). Therefore, bivariate cor-

relation was used to identify the factors that may influence indoor CO concentrations presented in

Table 1. A weak relationship exists between indoor and outdoor CO concentrations (r = 0.47),

which is generally normal. The value of the correlation coefficient (r) between the indoor and out-

door data can be used as indicator of the degree to which CO measured indoors is attributed to the

infiltration from outdoors. Chaloulakou and MaBVRoidis (2002) showed that the indoor peak-con-

centrations of CO are slightly dampened and lag behind outdoor peaks, thus suggesting that indoor

CO concentrations are not immediately affected by outdoor concentration changes, due to changes

in air exchange. Moreover, a positive correlation exists between indoor and outdoor CO concentra-

tions and PM10, PM2.5 and CO2 due to the same emission source. The PM–CO correlations observed

in this study are consistent with the findings of Dominick et al. (2012). The indoor CO concentra-

tion was found to be negatively correlated with indoor temperature and relative humidity. In natu-

rally ventilated buildings, the high building ventilation rate promptly brings indoor humidity to the

65

same level encountered outside. Thus, a negative correlation between humidity and CO infiltration

and/or build-up inside the building is expected. Furthermore, a negative correlation exists between

indoor and outdoor CO concentration and both ventilation rate and wind speed because low wind

speeds favor the accumulation of pollutants (low wind speeds are also related to stable atmospheric

conditions).

Table 1 Correlation coefficients (dimensionless) between indoor and outdoor CO concentrations and

meteorological Parameters.

Parameters CO (Indoor) CO (Outdoor)

CO (indoor) 1 0.47*

CO (outdoor) 0.47* 1

PM2.5(indoor) 0.38* 0.23*

PM2.5 (outdoor) 0.34* 0.19*

PM10 (indoor) 0.15* 0.13*

PM10 (outdoor) 0.29* 0.37*

CO2 (indoor) 0.26* 0.31*

CO2 (outdoor) 0.42* 0.40*

Temp (indoor) -0.41* -0.39*

Temp (outdoor) -0.42* -.34*

RH (indoor) -0.08* -0.02

RH (outdoor) -0.06 -0.01

VR -0.21* -0.16*

WS -0.10* -0.09*

*Correlation is significant at the 0.01 level (2-

tailed).

3.3 Principal component analysis

Sensitivity analysis, using FSR technique, was undertaken so as to examine the level of

importance for 15 variables i.e. outdoor CO, indoor and outdoor PM10, indoor and outdoor PM2.5,

indoor and outdoor CO2, temperature, relative humidity, dew point temperature, ventilation rate

,and wind speed. The standardised regression determination coefficient (R2) values were used to

estimate the relationship between indoor CO and these variables. The results show that 8 variables

were identified before the extraction using PCA as shown in Figure 2. After the extraction was

applied, three factors were considered as the principal component based on eigenvalues of more

than 1. PCA procedures was followed by a varimax rotation to maximize the loading of a predictor

variable and the higher loading variable with absolute values greater than 40% were selected for

66

the principal component interpretation (Abdul-Wahab et al., 2005). The factor loading values after

rotation are very important in interpretation of PCA results. The factor loadings correlate the factors

and the variables, the higher the factor loading, the more variable contributes to the variation of the

PC (Ul-Saufie et al., 2013). The eigenvalues for all linear components after rotation are shown in

Table 2. The cumulative variance of the principal components is 74.06%. The first PC explains 39.

98% of the total variation in the data set, which indicates a heavy load on relative humidity, tem-

perature dew point, ventilation rate, wind speed and outdoor CO. The second PC, which accounts

for approximately 20.96% of the total variation, loads heavily on indoor and outdoor PM10 and

PM2.5. Among the principal components, the third account for approximately 13.12% and load

heavily on outdoor and indoor temperature.

3.4 Multiple lin ear regression (MLR)

The stepwise MLR models of indoor CO prediction using the original parameters and PCs as the

inputs were conducted with regression assumptions approximately satisfied (Table 3). The four

goodness of fit measures showed that the residual distributions were approximately normal, with

zero means and no detectable serial and the correlation coefficients of the regressions were all

highly statistically significant (P< 0.01) (Abdul-Wahab et al., 2005).

Table 2 Rotated principal component loadings matrix

Component

1 2 3

CO(outdoor) -0.60

RH (indoor) -0.86

RH(outdoor) -0.80

VR 0.83

WS 0.82

CO2(indoor) 0.62

TDP(indoor) 0.90

TDP(outdoor) 0.90

PM2.5(indoor) 0.77

PM10(indoor) 0.77

PM10(outdoor ) 0.75

PM2.5(outdoor) 0.73

Temp(indoor) 0.93

Temp(outdoor) 0.93

Eigenvalue 6.00 3.14 1.97

% of Variance 39.98 20.96 13.12

Cumulative % 39.98 60.94 74.06

67

The result showed that the developed models did not encounter multicollinearity problems as the

VIF was less than 3.0. In addition, the tolerance values for the variables in MLR model are higher

than 0.3. In accordance with the findings of Field et al. (2009), the tolerance value must be smaller

than 0.1 to indicate a multicollinearity problem. However, DW may indicate slightly positive auto-

correlation problems in the model because these values ranged from 1.966 to 1.961. The developed

MLR and PCR models were also assessed using the coefficient of determination (R2), which was

used as an indicator of the ability of the selected variables to explain the variations in indoor CO

concentration (Abdul-Wahab et al., 2005).

Table 3 Summary models for indoor CO concentration predictions based on original parameters and PCA

as inputs.

Method Models R2 Range of VIF Durbin-

Watson

MLR CO (indoor) = 0.07 +1.07CO

(outdoor) + 0.07CO2(outdoor)

+0.06PM10(outdoor)-

0.10Temp(outdoor)

0.72 1.07-2.32 1.96

PCR CO (indoor) = 0.08+0.88PC1

+0.19PC2-0.03PC3

0.73 1.00 1.96

As presented in Table 3 when the four best variables are fitted to the CO data, the value of the R2

is approximately (0.72). Thus, approximately 72% of the variation in the indoor CO concentrations

can be explained by the four variables, as listed in the table. Meanwhile, the usage of PCs as the

inputs in MLR could improve the efficiency of the model to explain the variations in CO concen-

trations. During these time periods, R2 values for PCR was approximately (0.73), as reported in

Table 4. Therefore, approximately 73% of the variation in the indoor CO concentrations can be

explained by the three independent components.

In this study, the applicability of the developed models for predicting the indoor CO concentration

variations was assessed using hourly average monitoring records from validation data set. The per-

formance levels of the validation MLR and PCR models indicated that relatively strong relation-

ships were obtained between the observed and predicted values and R2 for these models was almost

0.87. Thus, R2 can explain approximately 87% of the variation in the indoor CO by using both

models. In addition to that, the values of RMSE ranged from 0.027 % for PCR model to 0.029% for

68

MLR model. By comparing the performance of the two models, even the R2 values for the both

models were comparable PCR model produced the lowest RMSE comparing with MLR.

3.5 Comparisons

Five performance indicators were used to evaluate and compare between the two models (MLR

and PCR) used to predict indoor CO as shown in Table 4. For a good model NAE and RMSE value

should approach zero, while PA, R2 and IA should be closer to one. The results suggest that the

four models are effective forecasting tools and hence can be applicable for short-term forecasting

of indoor CO level. PCA didn’t improve the accuracy measures of MLR model comparing with

PCR model where the results of PA, R2 and IA were almost comparable with MLR but PCA

reduced the prediction error in PCR comparing with MLR as calculated by the percentage

differences as much as 13.0% and 7.14% for NAE and RMSE, respectively. Thus, it can be

recommended that PCR could be a new promising methodology instead of a MLR to predict IAQ

in naturally ventilated buildings.

Table 4 Performance indicator for ANN models vs. MLR models for indoor CO

Model NAE RMSE R2 PA IA

MLR 0.213 0.029 0.869 0.937 0.965

PCR 0.187 0.027 0.870 0.937 0.961

4. CONCLUSION

Previously, MLR and PCR methods were used effectively to study air pollution and

meteorological records. In this study the capability of these techniques to predict indoor CO

concentrations in natural ventilated schools located in Gaza Strip, Palestine was employed. To raise

the efficiency of MLR, FSR method was used to select the key input variables for the optimal

structure of the models. PCA didn’t improve the accuracy measures of MLR model comparing with

PCR model where the results of PA, R2 and IA were almost comparable with MLR but PCA

reduced the prediction error in PCR comparing with MLR as calculated by the percentage

differences as much as 13.0% and 7.14% for NAE and RMSE, respectively. Overall, it was found

that the two models i.e. MLR and PCR are effective forecasting tools and hence can be applicable

for short-term forecasting of indoor CO level.

69

References

2. Abdul-Wahab, S. A., Bakheit, C. S. and Al-Alawi, S. M. (2005). Principal component and multiple regression analysis in modelling of ground-level ozone and factors affecting its concentrations. Environmental Modelling and Software, 20(10), 1263-1271.

3. Adar, S. D., Davey, M., Sullivan, J. R., Compher, M., Szpiro, A. and Sally Liu, L.-J. (2008).

Predicting airborne particle levels aboard Washington State school buses. Atmospheric Environment, 42(33), 7590-7599.

4. Agirre-Basurko, E., Ibarra-Berastegi, G. and Madariaga, I. (2006). Regression and multilayer

perceptron-based models to forecast hourly O3 and NO2 levels in the Bilbao area. Environmental Modelling and Software, 21(4), 430-446.

5. Al-Alawi, S. M., Abdul-Wahab, S. A. and Bakheit, C. S. (2008). Combining principal component

regression and artificial neural networks for more accurate predictions of ground-level ozone. Environmental Modelling & Software, 23(4), 396-403.

6. ATSDR (2012). Toxicological profile for CO. 1st ed.: Agency for Toxic Substances and Disease

Registry: U.S. Department of Health and Human Services. Atlanta

7. Blondeau, P., Iordache, V., Poupard, O., Genin, D. and Allard, F. (2005). Relationship between outdoor and indoor air quality in eight French schools. Indoor Air, 15(1), 2-12.

8. Braniš, M. and Šafránek, J. (2011). Characterization of coarse particulate matter in school gyms.

Environmental Research, 111(4), 485-491.

9. Cakmak, S., Dales, R. E. and Judek, S. (2006). Do gender, education, and income modify the effect of air pollution gases on cardiac disease? Journal of Occupational and Environmental Medicine, 48(1), 89-94.

10. Chaloulakou, A., Mavroidis, I. and Duci, A. (2003). Indoor and outdoor carbon monoxide

concentration relationships at different microenvironments in the Athens area. Chemosphere, 52(6), 1007-1019.

11. Chaloulakou, A., MaBVRoidis, I., 2002. "Comparison of indoor and outdoor concentrations of co

at a public school. Evaluation of an indoor air quality. Atmospheric Environment, 36 (2002) 1769–1781.

12. Chaloulakou, A., Mavroidis, I., Spyrellis, N. and (2001) In 7th International Conference on

Environmental Science and Technology Ermoupolis, Syros island, Greece.

13. Currie, J., Hanushek, E. A., Kahn, E. M., Neidell, M. and Rivkin, S. G. (2009). Does pollution increase school absences? The Review of Economics and Statistics, 91(4), 682-694.

14. Dominick, D., Juahir, H., Latif, M. T., Zain, S. M. and Aris, A. Z. (2012). Spatial assessment of

air quality patterns in Malaysia using multivariate analysis. Atmospheric Environment, 60(1), 172-181.

15. Elbayoumi, M., Ramli, N., Md Yusof, N. and Al Madhoun, W. (2014a). The effect of seasonal

variation on indoor and outdoor carbon monoxide concentrations in Eastern Mediterranean climate Atmospheric Pollution Research 5(2),315-324.

70

16. Elbayoumi, M., Ramli, N. A., Md Yusof, N. F. F., Yahaya, A. S. B., Al Madhoun, W. and Ul-Saufie, A. Z. (2014b). Multivariate methods for indoor PM10 and PM2.5 modelling in naturally ventilated schools buildings. Atmospheric Environment, 94, 11-21.

17. Gervasi, O. (2008). Computational Sicience and Its Applications-ICCSA 2008. Springer. Italy

18. Goniewicz, M. Ł., Czogała, J., Kośmider, L., Koszowski, B., Zielińska-Danch, W. and Sobczak,

A. (2009). Exposure to carbon monoxide from second-hand tobacco smoke in Polish pubs. Central European Journal of Public Health, 17(4), 220.

19. Gvozdić, V., Kovač-Andrić, E. and Brana, J. (2011). Influence of Meteorological Factors NO2,

SO2, CO and PM10 on the Concentration of O3 in the Urban Atmosphere of Eastern Croatia. Environmental modeling & assessment, 16(5), 491-501.

20. Hal. (2012). HAL-HPC300 user manual. [Online] Available from:

http://www.haltechnologies.com/Docs/HAL-HPC300%20User%20Manual.pdf [Accessed 02/10/2012].

21. HPA. (2009). Carbon monoxide:General information. [Online] Available

from:www.hpa.org.uk/webc/hpawebfile/hpaweb_c/1202487036268 [Accessed 29/11/2013].

22. Jef, H., Clemensa, M., Gerwinb, D., Fransb, F. and Olivierc, B. (2005). A neural network forecast for daily average PM10 concentrations in Belgium. Atmospheric Environment, 39(18), 3279-3289.

23. Kam, W, Cheung, K, Daher, N, Sioutas, C. 2011 “Particulate matter (PM) concentrations in un-

derground and ground-level rail systems of the Los Angeles metro,” Atmospheric Environmen, 45:1506-1516.

24. Koçak, M., Kubilay, N., Tugrul, S. and Mihalopoulos, N. (2010). Atmospheric nutrient inputs to

the northern levantine basin from a long-term observation: sources and comparison with riverine inputs. Biogeosciences, 7(12), 4037-4050.

25. Kovač-Andrić, E., Brana, J. and Gvozdić, V. (2009). Impact of meteorological factors on ozone

concentrations modelled by time series analysis and multivariate statistical methods. Ecological Informatics, 4(2), 117-122.

26. Kulshreshtha, P. and Khare, M. (2011). Indoor exploratory analysis of gaseous pollutants and

respirable particulate matter at residential homes of Delhi, India. Atmospheric Pollution Research, 2(3), 337-350.

27. Liao, D., Duan, Y., Whitsel, E. A., Zheng, Z.-j., Heiss, G., Chinchilli, V. M. and Lin, H.-M.

(2004). Association of higher levels of ambient criteria pollutants with impaired cardiac autonomic control: a population-based study. American Journal of Epidemiology, 159(8), 768-777.

28. Marković, D. M., Marković, D. A., Jovanović, A., Lazić, L. and Mijić, Z. (2008). Determination

of O3, NO2, SO2, CO and PM10 measured in Belgrade urban area. Environmental Monitoring and Assessment, 145(1-3), 349-359.

29. Özbay, B., Keskin, G. A., Doğruparmak, Ş. Ç. and Ayberk, S. (2011). Multivariate methods for

ground-level ozone modeling. Atmospheric Research, 102(1), 57-65.

71

30. PCBS (2012). Statistical Yearbook of Palestine Report, Palestinian Central Bureau of Statistics (PCBS), Ramalla- Palestine.

31. PMD. (2012). Climate Bulletine. [Online] Available from:

http://www.pmd.ps/ar/mna5palestine.php [Accessed 01/10/2012].

32. Poupard, O., Blondeau, P., Iordache, V. and Allard, F. (2005). Statistical analysis of parameters influencing the relationship between outdoor and indoor air quality in schools. Atmospheric Environment, 39(11), 2071-2080.

33. Raub, J. A. and Benignus, V. A. (2002). Carbon monoxide and the nervous system. Neuroscience

& Biobehavioral Reviews, 26(8), 925-940.

34. Slaughter, J. C., Lumley, T., Sheppard, L., Koenig, J. Q. and Shapiro, G. G. (2003). Effects of ambient air pollution on symptom severity and medication use in children with asthma. Annals of Allergy, Asthma and Immunology, 91(4), 346-353.

35. Sofuoglu, S. C., Sofuoglu, A., Birgili, S. and Tayfur, G. (2006). Forecasting ambient air SO2

concentrations using artificial neural networks. Energy Sources, Part B, 1(2), 127-136.

36. Ul-Saufie, A., Yahya, A. and Ramli, N. (2010). Improving multiple linear regression model using principal component analysis for predicting PM10 concentration in Seberang Prai, Pulau Pinang. International Journal of Environmental Sciences, 2(2), 403-408.

37. Ul-Saufie, A. Z., Yahaya, A. S., Ramli, N. A., Rosaida, N. and Hamid, H. A. (2013). Future daily

PM 10 concentrations prediction by combining regression models and feedforward backpropagation models with principle component analysis (PCA). Atmospheric Environment, 77, 621-630.

38. USEPA. (2013). An Introduction to Indoor Air Quality (IAQ), Carbon Monoxide (CO). [Online]

Available from: http://www.epa.gov/iaq/co.html [Accessed 12/03/2014].

39. WHO (2005). WHO Air Quality Guidelines Global Update 2005, World Health Organization (WHO), Bonn, Germany.

40. WHO (2011). Methods for monitoring indoor air quality in schools World Health Organization

(WHO) Bonn. Germany

41. Wilks, D. S. (2011). Statistical methods in the atmospheric sciences. Academic press

42. Yeniay, O. and Goktas, A. (2002). A comparison of partial least squares regression with other prediction methods. Hacettepe Journal of Mathematics and Statistics, 31(99), 99-111.

72

Evaluation of Indoor Air Quality at Engineering Campus Li-brary at the University Sains Malaysia

Maher Elbayoumi1, Nor Azam Ramli2, Noor Faizah Md Yusof2, Norrimi Rosaida Awang2, Nazatul Syadia Zainordin2, Maisarah Sulaiman2, Syabiha Shith2, Teh Nur

Amalina Zaki2 Received Feb 2018; accepted April 2018

Abstract: The indoor air quality (IAQ) in micro-environments is extremely important

due to its impact on health and productivity of students. This study presents the findings

of indoor air quality (IAQ) investigations in engineering campus library at the University

Sains Malaysia. Four levels of the library were investigated during May 2015.

Measurements were carried out by using electrochemical analyzer and hand-held

particulate matter (PM) instrument. The results showed that Formaldehyde, carbon

monoxide (CO), Ozone, and total volatile organic compounds (TVOC) levels were found

below Malaysian guideline values. Temperature ranged between 18.9–28.4 °C and relative

humidity (RH) ranged between 49%–73% with an average of 62.2%. The indoor

concentration PM2.5 level was 5.73±4.93μg/m3. All the monitoring pollutants’ levels were

found below Malaysian guideline values in the ground floor, and the 1st and 2nd floors.

However, several exceedances occurred in the 3rd floor. Inadequate introduction and/or

distribution of fresh air may be the main problem in the left side of the building and in the

3rd floor. Thus, evaluation of the HVAC (Heating, Ventilation, and Air Conditioning) system

should be examined.

Keywords: Indoor Air Quality, Thermal Comfort, HVAC, Ventilation, Student Health.

73

1. Introduction

Indoor air pollution has been ranked as the second biggest environmental provider to ill health

after unsafe water and sanitation (Keraka et al., 2013; Salem 2017) . Considering people spend an

estimated 90% of their time indoors, the air quality is an important factor related to occupant comfort

and health (USEPA, 2012; Salem 2017) . The most common health problems that indoor pollutants

may contribute are sick building syndrome (SBS) and building-related symptoms (BRS). The qual-

ity of indoor air has emerged into a relevant field of study and has received significant public atten-

tion since the energy crisis experienced in the early 1970s.

Various trade organizations and public agencies have developed recommendations and regulations

to minimize exposure to air contaminants; however, most of these exposure levels are established

for industrial environments. Mindful that non-industrial facilities do not use chemicals in frequency

and volume that is common at industrial establishments, the concentrations of indoor air contami-

nants in the former facilities are “rarely present in levels known to be harmful” (Conrad and Soule,

1997). Despite the media attention and academic research focused on chronic exposures to relatively

low concentrations of indoor contaminants, the adverse effects of poor indoor air quality (IAQ) at

non-industrial facilities remains ambiguous to the environmental and medical professional (Salem

2017).

Malaysia currently is experiencing rapid urbanization and economic growth due to the develop-

ment of industrial estates. The clear phenomenon of rural to urban migration has brought as a con-

sequence greater emissions into the atmosphere, which has predominantly been produced by the

increase in traffic. In addition, the expansion of suburbs into closer proximity with industrial plants

in certain areas has led to the problem of air pollution (Azmi et al., 2010) . As a tropical country,

the lower ventilation rates combining with increasing the outdoor pollutants may affect the indoor

air quality. Further, the high humidity and high temperatures experienced increase the risk of ther-

mal discomfort and moisture problems indoor (Hamimah et al., 2010).

Several IAQ studies in Malaysia reported that IAQ, thermal comfort and SBS has become a com-

mon issue in Malaysia buildings (Mustapha et al., 2008; Juliana et al., 2009; Makhtar et al., 2010;

Kamaruzzaman and Razak, 2011; Norhidayah et al., 2013). Moreover, till now no mandatory regu-

lations to be advised on IAQ in school buildings were established, where the regulations and stand-

ards have been set to industrial building during the last 10 years (DOSH, 2010; PWDM, 2013). The

purpose of this study is to examine indoor air quality (IAQ) levels, and thermal comfort within the

four floor of the University Sains Malaysia’s (USM) Engineering library.

74

2. Methodology

Overview

An industrial hygiene air sampling investigation has been conducted in the library in the USM’s

Engineering campus. Data collected during this investigation included fine particulate matter

(PM2.5), ground level ozone (O3), volatile organic compounds (TVOC), formaldehyde (CH2O), CO,

CO2, temperature (Temp), relative humidity (RH), and dew point temperature (Tdp). All monitoring

parameters were collected during one week, i.e. from 23-April 2015 through 30-April 2015.

Walk-through Survey

A walk-through survey was utilized to document the materials, design, condition and cleanliness

of the library, and the heating, ventilation, and air conditioning (HVAC) system. Data were

collected to characterize the materials and conditions of the ceiling, floor, interior walls, exterior

walls, HVAC equipment, library contents, environmental modifiers, and the indoor environment.

Among the specific items inspected were water stains, mould, air fresheners, pesticides, odours,

general cleanliness, and lighting quality.

Sampling Sites

Sampling was performed in the building during the hours of normal operation. The library building

utilizes HVAC unit to circulate the outdoor air with indoor air. In general, natural ventilation was

limited as windows and doors were closed throughout the study. Four floors of the library building

were selected to be monitored. In each floor two locations were chosen from the right and left sides

of the building. For every selected site in every floor 8 hours of continuous sampling was occurred.

Equipment and Supplies

A summary of the sampling equipment’s is provided in Table 1. The mass concentration of

particles (PM2.5) has been monitored using E-bam. The monitor performs particulate size

measurements by using laser light scattering method. GrayWolf IAQ monitor Model IQ-610 were

used for TVOC, O3, CO, CO2, Temp, RH and Tdp measurements. For formaldehyde measurement

was used Formaldemeter htV-m meter.

75

Table 1 Specifications of equipment’s used for data collection

Data

Collected

Range Company Model Uncertainty /

Detection Limit

Method

PM2.5 Size Range:

0.1-100

μm

Met One

Inc.

E-BAM-

9800 REV L

0.003 mg/m3 or 2%

Reading

90° light scattering,

laser diode with flow

rate 16.7 l/m

CO 0.0-500

ppm

GrayWolf

Sensing

Solutions

IAQ

monitor

Model IQ-

610

±2ppm <50ppm,

±3%rdg >50ppm

Electro-chemical

TVOC 5-20000

ppb

1 ppb Photo-Ionization

Detection (PID)

O3 - - Electro-chemical

CO2 0~10000

ppm

Accuracy: ± 3% rdg

± 50 ppm

NDIR (Non-

dispersive Infrared)

Temp -25° to

+70°C

±0.3°C Platinum

temperature (resistive

element)

RH 0 to 100 % ±2%RH <80%RH

(±3%RH>80%RH

Electrostatic

capacity

CH2O 0-10 ppm PPM

Technology

Formaldem

eter htV-m

Accuracy 10% at

2ppm

Electro-chemical

Sample Collection

Sampling was conducted in the selected locations during the study activities by using previous

mentioned samplers. The samplers were placed at least 1.5 m height from the floor as shown in

Figure 1. Samples were collected for 1 min over 8 hours.

Data Analysis

The data were analyzed using descriptive statistics and parametric statistical tests. The mean,

median, minimum, maximum and standard deviation of the data were determined. To compare the

between locations and floor data, the parametric t test and ANOVA were used, with a significance

value of 0.05, to determine if a statistically significant difference existed.

3. Results and Discussion

Overview

This study was designed to measure and evaluate the concentration of fine particulates, TVOC,

O3, CO, CO2, as well as the temperature and relative humidity within the investigated area at

the four floor’s library. The airborne contaminants are ubiquitous in the environment, and

therefore are expected at some measurable concentration in the indoor air. Temperature and

relative humidity were included in the study due to their influence on the building occupant’s

comfort level and overall perception of the IAQ.

76

Summary of Results

More than 3,800 data points were collected during this study. The data were analyzed to identify

the mean, median, minimum and maximum measurement and standard deviation for each

contaminant and comfort parameter. Summary statistics from the study are organized in Table 2.

. Figure 1 indoor air sampling in the USM’s Engineering library

Table 2. Descriptive Statistics of measured environmental parameters

Parameter Minimum Maximum Mean Std. Deviation

TVOC (ppm) 0.300 1.720 0.180 0.310

CO2 (ppm) 372.50 517.93 436.77 32.19

O3 (ppm) 0.00 0.00 0.00 0.00

CO (ppm) 0.00 1.36 0.59 0.32

CH2OH (ppm) 0.01 0.04 0.02 0.01

PM2.5 (μg/m3) 0.00 22.00 5.73 4.93

Temp (º c) 18.90 28.40 23.74 2.21

RH (%) 49.00 73.00 62.26 5.28

Tdp (º c) 12.28 21.92 15.60 2.35

Particulate Matter (PM2.5)

Particulate matter, or PM, refers to a mixture of solid and liquid atmospheric particles with an

aerodynamic diameter (d10) less than 10 micrometres. It arises mainly from anthropogenic sources,

such as fossil fuel combustion by electric utilities and motor vehicles, wood burning, and the

smelting or other processing of metals. Several population based studies have established a strong

77

correlation between exposures to particulate matter (PM) and increasing rates of mortality,

morbidity, respiratory and cardiovascular problems especially among children. The average

concentration of indoor PM2.5 during the study period was 5.73±4.93μg/m3 which is below the

values set by WHO guideline (25 μg/m3 for PM2.5). As illustrated in Figure 2 the trend of PM2.5

concentrations was the same during the sampling hours. When comparing the mean PM2.5

concentrations by the floors, a statistical difference was noted only between the 3rd floor (8.13

μg/m3) and 2nd, 1st and ground floors (6.19 μg/m3, 5.69 μg/m3 and 2.94 μg/m3, respectively) as shown

in Table 3. In addition by comparing between locations in the same floor (left and right sides), the

t-test showed the mean PM2.5 concentration was statistically different between locations only in 3rd

floor (p<0.001).

.

Figure 2 Temporal variation of indoor PM2.5

Table 3 Analysis of variance (ANOVA)

Sum of

Squares

Mea

n

Squar

e

F S

ig

.

Between

Groups

219.92 73.30 3

.

3

5

0

.0

2 Within

Groups

1310.56 21.84

Total 1530.48

78

Carbon Monoxide (CO)

Carbon monoxide is a by-product of incomplete combustion of organic matter (e.g, gasoline, wood

and tobacco). Exposure to carbon monoxide can produce immediate and acute health effects.

Several air quality standards have been established to address carbon monoxide and prevent

symptoms from exposure to these substances. The average indoor and outdoor CO concentration

for the library during the study period was very small and negligible (0.59±0.32 ppm). This value

is below the Malaysian Code of Practice (10 ppm) (DOSH 2010). This lower concentration may be

due to the University’s location as a suburban area, where CO is considered as a scale pollutant

which generated by road traffic. The natural ventilation was limited as windows and doors were

closed throughout the office hours of the library. As illustrated in Figure 3, CO concentrations was

slightly high in the morning (8:00-9:00) then decreased in the 1st ,2nd, and 3rd floors. When

comparing the mean CO concentrations by the floors, a statistical difference was noted only between

the 2nd, 3rd floors and 1st and ground floors. In addition by comparing between locations in the

same floor (left and right sides), the t-test showed the mean CO concentration in the left side of the

library was statistically different from the right side in all floors (p < 0.001).

Volatile Organic Compounds (TVOC)

Indoor air concentrations can be greatly impacted by the use of products containing volatile

organic compounds (VOCs). VOCs are carbon-containing substances that have the ability to

evaporate at room temperature. Frequently, exposure to low levels of total VOCs (TVOCs) may

produce eye, nose, throat and/or respiratory irritation in some sensitive individuals. Table 2 shows

the daily 8-hours average indoor-outdoor concentration levels of TVOC in the monitoring library.

The indoor concentration of TVOC ranged from 0.003 ppm to 1.72 ppm. Moreover, TVOC

concentrations indoor and outdoor were below the values set by the Malaysian Department of Safety

and Health for TVOC (3 ppm). As illustrated in Figure 4 TVOC concentrations was low and stable

during the monitoring hours. A unimodal pattern displayed with peaks at 1:00 PM in the 3 rd floor

then decreased. When comparing the mean TVOC concentrations by floor, a statistical difference

was noted only between the 3rd floor and the remaining floors. In addition by comparing between

locations in the same floor (left and right sides), the t-test showed the mean TVOC concentration in

the left side of the library was statistically different from the right side in all floors (p<0.001).

79

(a)

(b)

Figure 3(a) Temporal variation and (b) indoor CO concentration in library building

Formaldehyde (CH2OH)

The average level of formaldehyde during the sampling period was 0.02 ±0.01, which is below

the values set by the Malaysian Department of Safety and Health for CH2OH (0.1 ppm). As shown

in Figure 5 the concentrations in the fourth floor were small and stable.

80

Figure 4 Temporal variation of indoor TVOC

Ozone (O3)

It was assumed that the likely source of indoor O3 would be attributed to the presence of copy

machine. In general, the frequency of machine operation was minimal and most samples were

collected while the machine was idle. Regardless of whether the copy machine was in use, the indoor

concentration of O3 in the library was not detected during the study’s period. Therefore, no further

analyses were conducted.

Figure 5 Temporal variation of indoor formaldehyde

81

Carbon Dioxide (CO2)

Of the all measurements collected, a narrow range of CO2 levels were seen. The concentrations for

the four locations ranged from 372.5 ppm to 517.9 ppm. The mean CO2 reading for each location

was well below ASHRAE’s recommended level of 1,000 ppm. As illustrated in Figure 6, the CO2

concentration decreased as the day progressed from 517 ppm to 372 ppm during monitoring hours.

The four floors also displayed statistical differences in mean CO2 concentration with higher

concentration in the ground floor. The area at the ground floor is adjacent to the entrance doors. The

periodic opening of the doors could influence the mixing of the indoor and outdoor air, and, therefore,

offset significant differences CO2 in this area. Moreover, comparing the right and lift sides detected

statistical difference in CO2 in all floors (p < 0.001) .

Figure 6 Temporal variation of indoor CO2

Thermal Comfort Parameters

The mean temperature for all four locations was 23.74ºC, ranging from 18.9 ºC to 28.40 ºC

with a standard deviation of 2.21ºC. The total measurements for relative humidity were logged

with a range from 49% to 73%. The mean relative humidity for all four locations was 62.2%.

Both temperature and relative humidity are within optimum comfort the levels 23 and 26 º C

and 55% and 70% for temperature and humidity, respectively, in the ground level, 1st and 2nd

floors and exceeded the guideline occasionally in the 3rd floor. Figure 7 display the temporal

variation of indoor temperature in the 4th floors. As shown in the figure the temperature in the

3rd floor (26.5ºC) was significantly higher than the remaining floors (p < 0.001). A comparison

of the temperature within the individual floors revealed statistical differences in the ground level,

82

and the 2nd and 3rd floors in the right and left sides. Additionally, in comparing individual

locations, RH was statistically different in the left side than all the right side in all floors except

in the 3rd floor.

Figure 7 Temporal variation of indoor temperature and relative humidity

4. CONCLUSION

It is generally accepted among environment health and medical professionals alike, that poor

ventilation as well as contaminated indoor air can lead to complaints from building occupants. The

purpose of this study was to assess and evaluate the variability of PM2.5, O3, CO2, CO, TVOC,

formaldehyde, relative humidity, and temperature in the USM’s Engineering library. The result

showed that all environmental pollutants were below the Malaysian guideline of indoor air quality.

However, parametric data analysis indicated statistical differences of the chemical contaminants

83

and thermal comfort variables between the right side andthe left side of the building, where the

left side has significant chemical contaminants and thermal comfort variables, comparing with the

right side which may be due to pressure differentials between the two sides of the building that

may account for influx of contaminants. Further, by comparing the measured pollutants among the

4th floor, the 3rd floor had higher pollutants concentrations and a higher temperature. Inadequate

introduction and/or distributions of fresh air may be the main problem. Investigators may need to

discuss the operation of the ventilation system with buildings’ engineers and perform ventilation

testing to determine proper fresh air intake. Measurements should be made under maximum and

minimum airflow conditions to determine the range of fresh air intake.

5. References 6. Azmi, Siti Zawiyah, Mohd Talib Latif, Aida Shafawati Ismail, Liew Juneng, and Abdul Aziz

Jemain. "Trend and Status of Air Quality at Three Different Monitoring Stations in the Klang Val-ley, Malaysia." Air Quality, Atmosphere and Health 3, no. 1 (2010): 53-64.

7. Conrad, RG, and RD Soule. "The Occupational Environment–It’s Evaluation and Control." Am Ind Hyg Assoc J (1997): 104-29.

8. DOSH. "Code of Practice on Indoor Air Quality." Department of Safety and Health: Ministry of Human Resources Malaysia, 2010.

9. Hamimah, Siti, D. Baba, and L. Abd.Mutalib. "Indoor Air Quality Issues for Non-Industrial Work Place." International Journal of Research and Reviews in Applied Sciences 5, no. 3 (2010): 235-44.

10. Juliana, J, O Norhafizalina, ZA Azman, and J Kamaruzaman. "Indoor Air Quality and Sick Build-ing Syndrome in Malaysian Buildings." Global Journal of Health Science 1, no. 2 (2009): P126.

11. Kamaruzzaman, SN, and RA Razak. "Measuring Indoor Air Quality Performance in Malaysian Government Kindergarten." Journal of Building Performance 2, no. 1 (2011): 70-79.

12. Keraka, Margaret, Carolyne Ochieng, Jacobus Engelbrecht, and Charles Hongoro. "Association between the Use of Biomass Fuels on Respiratory Health of Workers in Food Catering Enterprises in Nairobi Kenya." Pan African Medical Journal 15, no. 12 (2013).

13. Makhtar, NK, AR Ismail, N Jusoh, and AP Puvanasvaran. "Thermal Comfort in Technical School: Physical Measurement Approach." Paper presented at the National Conference in Mechanical En-gineering Research and Postgraduate Studies, Pahang, Malaysia, 2010.

14. Mustapha, Arniatul Aiza, Seti Mariam Ayop, Muhammad Kamal Ahmad, and Fadzida Ismail. "A Thermal Comfort Study in Naturally Ventilated School Building in Malaysia." Built Environment 5, no. 2 (2008): 66-82.

15. Norhidayah, A, Lee Chia-Kuang, MK Azhar, and S Nurulwahida. "Indoor Air Quality and Sick Building Syndrome in Three Selected Buildings." Procedia Engineering 53 (2013): 93-98.

16. PWDM. "Guidelines on Indoor Environmental Quality (Ieq) for Government Office Building." Malaysia: Public Works Department of Malaysia, 2013.

84

17. Salem, H.S. Indoor air pollution sources (particularly Skunk) and their impacts on health and the environment in the Occupied Palestinian Territories. (PP: 204-221). In M.F. Yassin (Ed.), "Pro-ceeding of Workshop on Air Quality in Hot Arid Climate (IAQHAC). Kuwait Institute for Scien-tific Research (KISR), Shuwaikh, Kuwait City, Kuwait, 3–4 April 2017.

18. USEPA. "Indoor Air Quality, Tools for Schools." Enviromnetal Protection Agency (EPA), http://www.epa.gov/iaq/schools/

85

Estimation of Temperature in Aluminum Plasma Using Laser

Induced Breakdown Spectroscopy

Shawqi A. M. Mansour *3

Received Jan 2018; accepted April 2018

Abstract: Plasma produced by a 1.064 µm pulsed with a pulse duration of 6ns focused onto a

pure aluminum solid sample (~99.99%) in air at atmospheric pressure is studied spectroscopy.

An Echelle spectrograph coupled with a gate intensified charge coupled detector is used to

record the plasma emissions. The plasma temperature was measured by time-resolved

spectroscopy of aluminum neutral atom line emissions in the time window of 1-5 µs, using Saha-

Boltzmann plot method. The aluminum neutral lines were found to suffer from optical thickness

over the entire delay times. Analytical relations were used and experimental procedures devised

for evaluation of the self-absorption coefficients of several Al-lines, which are important to get

reliable temperature measurements. The results shows that Al (I) lines have highest plasma

temperature of 1.427 eV before correction against self absorption, while revealed a lowest

temperature of 1.092 eV after correction.

Key words: Optical emission spectroscopy (OES), Electron temperature, Self-bsorption

effect (SA), 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 𝑛𝑚), LIBS.

3 Shawqi A. M. Mansour, Energy and Sustainable Environment Center, Palestine Technical College – Deir El Balah, Engineering Department-Deir El - Balah , Palestine. dr_shmansour@yahoo.com.

86

1. INTRODUCTION

Laser-induced breakdown spectroscopy is a powerful analytical technique with a wide range of

applications concerning both fundamental purposes and material analysis (Tognoni et al. 2002) ,

(Aragón et al. 2005), (Bengoechea et al. 2006), (Colón and Alonso-Medina 2006). LIBS

techniques is unique in that it may be used to chemically analyze rocks, glasses, metals, sand, teeth,

bones, powders, hazardous materials, liquids, plant, biological material, polymers, ceramics, etc.

(Thoron, 1988), (Amoruso et al., 1997). Aluminum metal as an abrasive material due to its

hardness and as refractory material due to its high melting point remained a subject of interest for

many researchers. (Lu et al., 1999) studied the aluminum plasma generated by an excimer (248nm)

in vacuum using an optical multichannel analyzer. (Colón et al., 1993) measured the Stark

broadening parameters of some of the All transition lines in the nitrogen atmosphere. (Body and

Shadwich , 2001) presented new instrumentation variation on LIBS and used it for the analysis of

coal by detecting AL, Si, Mg, Ca, Fe, Na, C, K and H as its key organic components. The

consequence of the ambient gas pressure on the spectral intensity of the zinc-aluminum alloy was

studied by (Kim et al., 1997) using a Nd: YAG laser at 1064 nm and reported that the spectral

intensity of the neutral aluminum line at 309.27 nm increases with the ambient pressure in air and

argon. (Handoco et al., 2006) reported the time-resolved emission spectroscopic investigations of

pulse laser-ablated plasma of ZrO2 and Al2O3.

In this paper, we shall adopt a straightforward procedure to calculate the self-absorption

coefficients of the plasma to the resonance lines at 308.21nm, 309..27nm, 394.40nm, and 396.15nm

of neutral aluminum via comparison of the electron densities evaluated using the Stark broadening

of the aluminum lines to the electron density as deduced from the optically thin 𝐻𝛼 − 𝑙𝑖𝑛𝑒

(Mansour, 2015). The induced plasma temperature was determined after the correction of

aluminum intensities lines against self-absorption effect.

1. DETERMINATION OF THE ELECTRON TEMPERATURE

Knowledge of the plasma temperature is vital to understand the dissociation, atomization,

ionization and excitation processes occurring in the plasma and helpful in utilizing the plasma to

maximize analytical potential of LIBS (McWhirter R. W. P., 1965). The temperature, T defined

as the particular form of energy and it can determine the equilibrium distribution of energy among

the different state of the particle assembled. It may happen that the equilibrium distributions exist

for one form of energy but not for another. Thermodynamic equilibrium will exist when all form of

energy distribution is described by the same temperature.

87

In the case of Local Thermodynamic Equilibrium (LTE), the excitation temperature, 𝑇𝑒𝑥𝑐. is equal

to temperature of electron, 𝑇𝑒 and 𝑇𝑖 temperature of heavy particle i.e., atom and ions, 𝑇𝑒𝑥𝑐. = 𝑇𝑒 =

𝑇𝑖 where, 𝑇𝑖 is the temperature describing the photon distribution. The escape of photon is

associated with spatial gradient in the plasma and to time-dependent regimes, so that LTE can be

established (Fujimoto, 2004), (Van Der Mullen, 1990), and (Capitelli et al., 2000).

In LIBS plasma the ionization degree is sufficiently high, this completion dominated by the elec-

tron, i.e. 𝑇𝑒𝑥𝑐. ~ 𝑇𝑖 and just a small perturbation that is usually be neglected can be expected from

the temperatures of electron and heavy particle (Cristoforetti et al., 2010). In typical LIBS plasma

only, neutral atom and singly charged ion are presents to a significant degree. Therefore, only neu-

tral and singly ionized particle will be considered. Under LTE condition, the population of the ex-

cited level for each species follows a Boltzmann distribution (Miziolek et al., 2006). The condition

of atomic and ionic state should be populated mainly by electron collision other than radiation, to

ensure it has high collision rate the electron density must be sufficient. The minimum limit for elec-

tron density 𝑛𝑒 is,

𝒏𝒆 = 𝟏. 𝟔 × 𝟏𝟎𝟏𝟐𝑻𝟏/𝟐(∆𝑬)𝟑 … … … … … . . … … … . . (𝟏)

where ΔE is the highest energy to hold the LTE condition, and T is the plasma temperature. This

limit is given by McWhirter criterion to fulfill during the first stage of plasma lifetime. This criterion

is necessary even though it insufficient for the condition (Miziolek et al., 2006). The excitation

temperature control the population of atomic and ionic energy level must be same as the ionization

temperature. It resolved the distribution of atom of the same element in the different ionization

stages. It describe in Saha equation where the neutral and singly ionized species of the same element

can be written as,

𝒏𝒆

𝒏𝑰𝑰

𝒏𝑰=

𝟐(𝟐𝝅𝒎𝒆 𝒌𝑻)𝟑𝟐

𝒉𝟑

𝟐𝑼𝑰𝑰(𝑻)

𝑼𝑰(𝑻) 𝒆−

𝑬𝒊𝒐𝒏𝒌𝑻 … … … … (𝟐)

Where, 𝑛𝑒 is the plasma electron density, 𝑛𝐼 and 𝑛𝐼𝐼 are the number densities of the neutral atomic

species and the single ionized species, respectively, Eion is the ionization potential of the neutral

species in its ground state, me is the electron mass, and h is Planck’s constant. In accurate

calculations, the ionization potential lowering factor 𝐸𝑖𝑜𝑛 should be taken into accounts for the

typical value being on the order of 0.1 eV. In the measurement of plasma temperature, many

methods have described it based on the absolute or relative line intensity (line pair ratio or

Boltzmann plot), the ratio line to the continuum intensity. The method was depending on the

88

experimental condition whether it is suitable or not (Bye and Scheeline, 2003). Boltzmann equation

is use to relate the population of an excited level to the total number density of the species in the

plasma. After the linearization, the formula of Boltzmann plot obtained was:

𝒍𝒏𝑰𝒊𝒋

𝒈𝒊 𝑨𝒊𝒋= 𝒍𝒏 (

𝒏𝑺(𝑻)

𝑼𝑺(𝑻)) −

𝑬𝒊

𝒌𝑻… … … … … … . . . . (𝟑)

The left hand side of the equation (3) versus 𝐸𝑖 was plotted and has a slope of -1/kT. The plasma

temperature can be calculated without ns and 𝑈𝑠(𝑇). The gradient usually gives negative slope. The

electron can be derived from the intensity ratio of the two lines corresponding to the different

ionization stages of the same element when the plasma is near to LTE condition. The formula of

Saha equation refer to the ratio of the total number densities of two ionization stages of the same

element. Most of the research is using Stark broadening method and use the line intensity ratio to

determine the electron density of plasma (Detalle et al., 2001), (Le Drogoff et al., 2010), (Lee et

al., 1992) and (Sabsabi et al., 1995).

2. DETERMINATION OF THE ELECTRON DENSITY

The electron density is an important parameter used to describe the plasma environment and it is

crucial for establishing its equilibrium status. It can be evaluated from the profile of the spectral

line emitted through a line of sight of laser-produced plasma, which is the result of several spectral

broadening and shift mechanisms (Griem, 1964) and (Konjevic et al., 2000). In the experimental

conditions of the present work, the main contribution to line widths arises from Stark effect where

the contributions of other mechanisms of broadening can be neglected. The FWHM of the spectral

line under study 𝛥𝜆1𝑙2 was determined by a Voigt fitting procedure. Hence, the electron density (in

𝑐𝑚−3) can be determined from the line width as.

∆𝜆12⁄ ≅ 2𝜔 (

𝑛𝑒

𝑁) … … … … … … … … … … … … … … (𝟒)

where, 𝜔 is the electron impact width parameter, Nr is the reference electron density which equal

to (1016𝑐𝑚−3) for neutral atoms and (1017𝑐𝑚−3) for singly charged ions (Konjevic, 2000) and

(Konjevic et al. , 2003). In the special case of the hydrogen 𝑛𝑒 (𝐻𝛼) − 𝑙𝑖𝑛𝑒 the electron density

can be related to the Lorentzian half width at the half of the maximum 𝛥𝜆1𝑙2 through the relation

(Griem, 1964),

𝐻𝛼 = 8.02 × 1012 (∆𝜆𝑆

𝛼12⁄

) 𝑐𝑚−3 … … … … … . . … . (𝟓)

Where, ∆𝜆𝑠 is the intrinsic full width at half of maximum (FWHM) of the spectral line in

Angstrom, and α1/2 is the half width of the reduced Stark profiles in Angstrom. Precise values of

𝛼1𝑙2 for the Balmer series can be found in (Konjevic et al., 2003), (Kepple and Griem, 1968) and

89

(Almen, 1987). However, when the laser beam is focused on the target, the ablation of the target

takes place, and due to the density gradient, the plasma rapidly expands.

3. SELF-ABSORPTİON ANALYSİS:

The self-absorption (SA) coefficient focused on line (𝜆𝑜) results from the expression is (Griem,

1974),

𝑆. 𝐴 =𝐼(𝜆𝑜)

𝐼𝑜(𝜆𝑜)=

[1 − 𝑒−𝑘(𝜆𝑜)𝑙]

[𝐾(𝜆0)𝑙]… … … … … … … (𝟔)

where, 𝐼(𝜆𝑜)is the relative intensity subject to self-absorption, 𝐼𝑜(𝜆𝑜) is the same as the former,

but without self-absorption, and 𝑘(𝜆𝑜)𝑙 = 𝜏(𝜆𝑜) is the plasma optical path until the center of the

line. On the other hand, the self-absorption coefficients of Al I- lines have been estimated from the

basic relation as described in (El Sherbini et al., 2005), (Mansour, 2015), (Mansour et al.,

2015), and (Mansour, (2017).

𝑆. 𝐴 = (∆𝜆

∆𝜆𝑜)

1𝛼⁄

= (𝑛𝑒 (𝑙𝑖𝑛𝑒)

𝑛𝑒(𝐻𝛼))

1𝛼⁄

… … … … . … … … (𝟕)

Where, 𝑛𝑒(line) is the electron density of the line which suffering from self-absorption effect,

𝑛𝑒 (𝐻𝛼) is the electron density of 𝐻𝛼 − 𝑙𝑖𝑛𝑒 free from self-absorption and 𝛼 = 0.56. Hence, the

researchers utilized Eq. (3) in order to calculate the amount of absorption (SA).

4. Experimental details

Figure 1 show the schematic diagram of LIBS experiment. The Q-Switched Nd:YAG laser

(Quintal, model Brilliant B) was operated at fundamental wavelength of 1064 nm, the repetition rate

of 10 Hz and pulse width of 6 ns, which capable of delivering 670 mJ at 1064 nm. An absolute

calibrated power-meter (Ophier, model 1z02165) was used for measuring a fraction of the laser light

reflected from a quartz beam splitter to monitor the incident laser energy. The laser beam was

focused on the target using convex lens of focal length 10cm. The sample was mounted on a three

dimensional sample stage, which was rotated to avoid the non-uniform pitting of the target. The

distance between the focusing lens and the sample was kept at 9.5cm, less than the focal length of

the lens to prevent any breakdown of the ambient air in the front of the target. The spectra were

obtained by averaging three single data of shots under identical experimental conditions. The laser

spot was measured at the target surface and gives a circle of diameter of 2mm because of the

deflagration effect and hence the laser energy per pulse of the order of 327.42mJ was measured at

the target surface. The radiations emitted by the plasma were collected by quarts optical fiber (with

90

a 25μm diameter) placed at right angle to the direction of the laser beam. The optical fiber was

connected with detection system consists of Echelle spectrograph (catalina, model SE 200) equipped

with a time gated ICCD Camera (1064×1064 pix with 13μm ×13μm pixel size at a binning mode

of 1×1 (type Andor , model iStar DH 737-18F). The wavelength scale was calibrated using a low

pressure Hg-lamp (Ocean optics). The instrumental bandwidth was measured from the FWHM of

the Hg -lines and was found on the average to be 0.12 ± 0.02 nm. Identification of the different lines

in the LIBS Spectrum was carried out using Spectrum Analyzer Software version 1.6. The

experimental setup including the Optical fiber was absolutely calibrated using a deuterium tungsten

halogen lamp (type Ocean optics, model DH 2000 Cal.).

Figure (1): The schematic diagram of the experiment

6. Results and Discussion

6.1. Plasma emission and spectral line analysis:

Figure 2 shows the plasma emission spectra of the aluminum sample. The resonance lines detected

were 308.21, 309.27, 394.40, and 396,15 nm have been used to infer the plasma temperature.

Assignments of these lines are taken from the NIST data base , and are listed with related details in

Table 1. We can observed clearly that the lines spectra are supper imposed on a large continuum

component. This continuum is mainly results from the free- free (Bremsstrahlung process) and the

free- bound transitions. This continuum should be removed before proceeding in the spectra line

shape analysis.

91

Figure (2): Shows the whole LIBS plasma emission spectra of sample aluminum for wavelength

1064 nm (red color), and the spectral lines lies under study ( solid black).

Table 1: Spectroscopic data of aluminum lines.

Transition Excitation

Energy (eV)

Statistical

Weight(g)

Transition

Probability

A(sec-1)

Wavelength

λ(nm)

Element

3s23p4s1p→3s23p21d 1.93 1 3.83

×108

281.61 Al II

2D3/2 p→3p2 2P1/2 4. 021 4 6.30

×107

308.206 Al I

3d 3D3/2→3p 2 P3/2 4. 022 5 4.93×107 309.27 Al I

4s 2S1/2 →3p 2P1/2 3.143 2 4.93×107 394.40 Al I

4s 2S1/2 →3p 2P3/2 3.143 2 9.8×107 396.15 Al I

6.2. Electron density determination and self-absorption analysis:

In this spectroscopic analysis, the resonance neutral Al lines at 308.21 nm, 309.27 nm, 394.40 nm

and 396.15 nm were selected for determining the plasma temperature Te. The isolated optically thin

hydrogen 𝐻𝛼 − 𝑙𝑖𝑛𝑒 at 656.27 nm appeared in the spectrum was used to determine the plasma elec-

tron density Ne (Figure 3) and to correct the Al (I) lines which contained some optical thickness.

The electron number density in the plume is determined using Eq. (4) for Al(II,I) lines and Eq.(5)

for hydrogen line at the wavelength 656.27 nm. It is evident that, in Figures(5-8),the deviation of

the measured electron density calculated from the Aluminum neutral lines with respect to that esti-

92

mated from the optically thin 𝐻𝛼 − 𝑙𝑖𝑛𝑒 indicates the existence of self-absorption while, the alumi-

num ionic line exhibits a good agreement with 𝐻𝛼 − 𝑙𝑖𝑛𝑒 reminiscent of free from self-absorption

as shown in Figure 4.

Figure (3): Shows the variation of the electron density of 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 nm) with laser energy.

Figure (4): Comparison of electron densities between the 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 nm) and Al II (281.61nm) at dif-

ferent laser energies.

93

Figure (5): Comparison of electron densities between the 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 nm) and Al I (308.21

nm) at different laser energies.

Figure (6): Comparison of electron densities between the 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 nm) and Al I (309.27

nm) at different laser energies.

Figure (7): Comparison of electron densities between the 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 nm) and Al I

(394.40 nm) at different laser energies.

94

Figure (8): Comparison of electron densities between the 𝐻𝛼 − 𝑙𝑖𝑛𝑒(656.27 nm) and Al I (396.15

nm) at different laser energies.

6.3. Validity of the Local Thermodynamic Equilibrium ( LTE):

To determine the electron temperature, the plasma must satisfy the equilibrium conditions, i.e. the

plasma must hold a state of local thermodynamic equilibrium during the observation window. In an

LTE plasma, the collisional excitation and de-excitation processes must dominate radiative

processes and this requires a minimum electron density. The lower limit of the electron density for

which the plasma will be verify Eq.(1), where ∆𝐸 (eV) is the difference between the upper and

lower states and T (eV) is the temperature. For the Al(I) 308.61 nm line transition, ∆𝐸 = 4.021 𝑒𝑉,

and the highest electron temperature observed was approximately 1.74𝑒𝑉. From eq. (1), a minimum

electron density of 1.093 × 1016 𝑐𝑚−3 is required for LTE to hold, which is much lower than

the N𝑒(1017𝑐𝑚−3) obtained in our experiments. Therefore, local thermodynamic equilibrium is

valid for the condition of the present plasma.

6.4. 6. d . Electron temperature determination (Te):

When evaluating the electron temperature using Saha-Boltzmann plot method and the electron

number density using spectral line broadening, it is important to correct the intensities of aluminum

resonance lines against self-absorption effect according to Eq.(7). On the other hand, Figure 9 shows

the Saha-Boltzmann plot for aluminum element. λ and I are the wavelength and the intensity of the

spectral lines. The temperature was obtained from the slope of the lines. The plasma temperature of

aluminum was 1.44eV before correction against self-absorption effect while became 1.09 eV after

the correction, which demonstrates the importance of treatment of self-absorption effect before

estimating the plasma temperature .The variation of temperature of Al laser produced plasma in

aluminum sample at different laser pulse energies is shown in ( Figure 9) below. This figure clearly

indicates that with the increase in the laser pulse energy the plasma temperature is also increased.

This due to the absorption and /or reflection of the laser photon by the plasma (Hafeez et al., 2015).

95

7. CONCLUSION:

A Q-switched laser of Nd: YAG (of 670 mJ per pulse) focused on an aluminum sample, at differ-

ent laser energies in air, has been used for the plasma diagnosis produced by laser (LPP). The Stark

broadening of the Hα − line Hydrogen (656.27 nm) has been used to estimate the electron density

of plasma obtained at different laser energies conditions. We have tried measured the electron den-

sities with the profile of the 281.62 nm line of Al II, and the profiles of the 308.21, 309.07, 394.40,

and 396,15 nm lines of Al I. The results indicate us that the ionic line doesn’t present self-absorption.

In contrast the aluminum resonance lines exhibit self-absorption and its usage is inadvisable in

plasma temperature estimation. The plasma temperature was calculated before self-absorption quan-

tification and was found to be around 16704 K, while became around 11,000 K after quantification.

Figure (9): Comparison of Boltzmann plots using Al (I) lines at 308.21 nm, 309.27 nm, 394.40 nm,

and 396.15 nm before SA correction and after SA correction

96

Figure (10): Variation of plasma temperature with laser energy after correction against Self-

absorption.

8. ACKNOWLEDGEMENTS

The researcher is thankful to the laboratory of lasers and new materials (LLNM) in Cairo

University in Egypt for the encouragement in terms of provision of time and moral support to carry

out the research work.

References

5. Almen M.V., (1987) Laser-beam interactions with Materials, Springer-Verlag, New York.

6. Amoruso, S., Armenante, M., Berardi, V., Bruzzese, R. and Spinelli, R., (1997) Absorption and Saturation Mechanism in Aluminum Laser Ablated Plasmas. Appl. Physics. A. 65: pp 265-271.

7. Aragón, Bengoechea, A. J., Aguilera, J.A., (2005) Asymmetric Stark broadening of the Fe I 538.34 nm emission line in a laser induced plasma, Spectrochim. Acta Part B 60 pp 897–904.

8. Bengoechea, J. J, Aguilera, A., Aragón, C., (2006) Application of laser-induced plasma spectros-copy to the measurement of Stark broadening parameters, Spectrochim. Acta Part B 61 pp 69–80.

9. Body, D. and Chadwick, B. L., (2001) Optimization of the spectral data processing in a LIBS, Rev. Sci. Instrum., Vol. 72, pp. 1625-1629.

10. Bye, C. A., and Scheeline, A., (2003) Laser-Induced Breakdown Spectroscopy (LIBS) Fundamentals and Applications. Appli. Spectrosc. 2022-2030.

11. Capitelli, M., Capitelli, F. and Eletskii, A., (2000) Non-Equilibrium and Equilibrium Problems in Laser-Induced Plasmas, Spectrochim. Acta Part B. 55: pp 559-574.

12. Colón, C. and Alonso-Medina A., (2006) Application of a laser produced plasma: ex-perimental Stark widths of single ionized lead lines, Spectrochim. Acta Part B 61 pp 856–863.

13. Colón, C., Htern, G., Verdugo, P., Ruiz, P., and Campos, J.,(1993) Measurements of the Stark broadening and shifts parameters for several ultraviolet lines of singly ionized aluminum, J. Appl. Phys., Vol. 73, no. 10, pp 4752-4758.

14. Cristoforetti, A. g., De Giacomo, A. B., Dell'aglio C. M., Legnaioli, A. E. S., Tognoni, A. V., Palleschi, A. N. and Omenetto, D. (2010) Local Thermodynamic Equilibrium in Laser-Induced Breakdown Spectroscopy: Beyond The Mcwhirter Criterion. Spectrochimica Acta Part B. 65: 86-95.

15. Detalle, V., Hόon, M., Sabsabi, M., and St-Onge, I., (2001) An evaluation of a commer-cial Echelle spectrometer with intensified charge-coupled device detector for materials analysis by laser-induced laser plasma spectroscopy, Spectrochim. Acta B, At. Spec-trosc., Vol.56, no. 6., pp. 1011-1025.

16. El Sherbini, A., El Sherbini, M. Th, M. Heagzy, H., Cristoforetti, G., Legnaioli, S.,Palleschi, V., Pardini, L., Salvetti, A., Tognoni, E., (2005) Evaluation of self-absorption of aluminum emission lines in laser-induced breakdown spectroscopy measurements, Spectrochim. Acta B 60, pp1573–1579.

17. Fujimoto, T., (2004) Plasma Spectroscopy. Clarendon Press, Oxford. 18. Griem H.R., (1964) Spectral Line Broadening by Plasmas, Academic Press, New York. 19. Griem, H.R., (1974) Spectral Line Broadening by Plasmas. Academic Press, New York. 20. Hafeez, S., Shaikh, N. M. and Baig M. A., (2008) Spectroscopic studies of Ca plasma generated

by the fundamental, second, and third harmonics of a Nd:YAG laser, Laser and Particle Beams 26 pp41.

97

21. Handoko, A. D., Lee, P. S., Lee, P., Mohanty S. R., and Rawat, R. S., (2006) Time resolved emis-sion spectroscopy, J. Phys.: Conf. Ser., Vol. 28, pp 100-104.

22. Kepple P., and Griem H.R., (1968) Improved Stark Profile Calculations for Hydrogen Lines: Hα, Hβ,Hδ and Hγ, Physical review, 173 pp 317-325.

23. Kim, D. E., Yoo, K. J, Park, H. K., Oh, k. J., and Kim, D. W., (1997) Quantitative analysis of aluminum impurities in Zinc alloy, Appl. Spectrosc., Vol. 51, no.1, pp 22-29.

24. Konjevic N., (2003) Plasma Broadening and Shifting of Non-Hydrogenic Spectral Lines: Present Status and Applications, Physics Reports, 316 pp 339- 401.

25. Konjevic N., Lesage A., Fuhr J. R. and Wiese W. L., (2000) Experimental Stark Width for Spec-tral Lines of Neutral and Ionized Atoms (A Critical Review of Selected Data For the Period 1989 through2000,Journal of Physical and Chemical Reference Data, 31 pp 819- 927.

26. Le Drogoff, B., Margot, J., Chacker, M., Sabsabi, M., Barthelemy, O., Johnston, T. W., Laville, S., Vidal, F., and von Kaenel, y., (2001) Temporal characterization of femtosecond laser, Spectro-chim. Acta B , At. Spectrosc., Vol. 6, no. 6, pp 987-1002.

27. Lee, Y. L., Sawan, S. P., Thiem, T. L., Teng, Y.-Y., and Sneddon, j., (1992) Interaction of a laser beam with metals, part II: Space-resolved studies of laser-ablated plasma emission, Appl. Spec-trosc., Vol. 46, no.3, pp. 436-441.

28. Lu, Y., Tao, z., and Hong, M., (1999) Characteristics of excimer laser induced plasma from an aluminum target, Jpn. J. Appl. Phys., Vol.38, no. 5A, pp 2958-2963.

29. Mansour S. A., (2015) Self-Absorption Effects on Electron Temperature Measurements Utilizing Laser Induced Breakdown Spectroscopy (LIBS)-Techniques, Optics and Photonics Journal, 5 pp79-90.

30. Mansour S. A., El Sherbini A. M., Allam S. H., El Sherbini Th. M., (2015) Optical Spectroscopic Studies of Silicon Plasma Produced by Fundamental (1064) of an Nd: YAG Laser, Australian Journal of Modern Physics,2 pp1-13.

31. Mansour S. A., El Sherbini A. M., Allam S. H., El Sherbini Th. M., (2017) Spectral Studies of Na-nosecond Laser Interaction with Aluminum-Based Alloy Containing Traces of Magnesium in Air, Israa University Journal of Applied Science,1 pp55-67.

32. McWhirter R. W. P, (1965) Plasma Diagnostic Techniques Pure and Applied Physics vol. 21eds Huddlestone R H and Leonard S. L. (New York: Academic Press) 201-264.

33. Miziolek, A. W., Palleschi, V. and Schechter, I., ( 2006) Laser-Induced Breakdown Spectroscopy (LIBS) Fundamentals and Applications. Cambridge University Press, United Kingdom, Britain.

34. NIST Atomic Spectra Database (ver. 5.4), http://physics.nist.gov/asd (2017). 35. Sabsabi, M., and Cielo, P., (1995) Quantitative analysis of aluminum alloys by laser-induced

breakdown spectroscopy , Appl. Spectrosc., Vol. 49, no. 4, pp. 499-507.Yalcin S., Crosley D. R., Smith G. P. and Faris G.W.(1999) Influence of Ambient Conditions on The Laser Air Spark, Ap-plied Physics: B Laser and Optics, 68 pp 121-130.

36. Thoron, A. P., (1988) Spectroscopic, 2nd ed. London, U.K.: Chapman & Hall. 37. Tognoni, E. Palleschi, V. Corsi, M. Cristoforetti, G., (2002) Quantitative microanalysis by laser-

induced breakdown spectroscopy: a review of the experimental approaches, Spectrochim. Acta Part B 57 pp1115–130.

38. Van Der Mullen, J. M. N., (1990) On the Atomic State Distribution Function in Inductively Cou-pled Plasmas -The Stage of Local Thermal Equilibrium and its Validity Region. Spectrochim Acta Part B. 45: 1-13.