Journal of Environmental

Science and Engineering A

Volume 1, Number 1, January 2012 (Serial Number 1)

David Publishing

David Publishing Company

www.davidpublishing.org
Publication Information:
Journal of Environmental Science and Engineering A (formerly parts of Journal of Environmental Science and
Engineering ISSN 1934-8932, USA) is published monthly in hard copy (ISSN 2162-5298) and online (ISSN
2162-5301) by David Publishing Company located at 9460 Telstar Ave Suite 5, EL Monte, CA 91731, USA.

Aims and Scope:

Journal of Environmental Science and Engineering A, a monthly professional academic journal, covers all sorts of
researches on environmental management and assessment, environmental monitoring, atmospheric environment,
aquatic environment and municipal solid waste, etc..

Editorial Board Members:

Dr. Bishnu Rajupreti (Nepal), Prof. Jianhua Wang (China), Prof. Mankolli Hysen (Albania), Dr. Jungkon Kim (South
Korea), Prof. Samira Ibrahim Korfali (Lebanon), Prof. Pradeep K. Naik (Bahrain), Dr. Ricardo García Mira (Spain), Dr.
Leucci Giovanni (Italy), Prof. Konstantinos C. Makris (Gonia Athinon & Nikou Xiouta), Prof. Kihong Park (South
Korea), Prof. Mukesh Sharma (India), Dr. Hesham Gehad Mohamed Ibrahim (Palestine), Dr. Jyoti Prakash Maity
(India), Dr. Giuseppe Mascolo (Italy), Dr. Satinder Kaur Brar (Canada), Dr. Jo-Ming Tseng (Taiwan), Associate Prof.
Muntean Edward Ioan (Romania).

Manuscripts and correspondence are invited for publication. You can submit your papers via Web Submission, or
E-mail to environmental@davidpublishing.org or info@davidpublishing.org. Submission guidelines and Web
Submission system are available at http://www.davidpublishing.org.

Editorial Office:
9460 Telstar Ave Suite 5, EL Monte, CA 91731, USA
Tel: 1-323-9847526, 1-302-5977046
Fax: 1-323-9847374
E-mail: environmental@davidpublishing.org; info@davidpublishing.org.

Copyright©2012 by David Publishing Company and individual contributors. All rights reserved. David Publishing
Company holds the exclusive copyright of all the contents of this journal. In accordance with the international
convention, no part of this journal may be reproduced or transmitted by any media or publishing organs (including
various websites) without the written permission of the copyright holder. Otherwise, any conduct would be considered
as the violation of the copyright. The contents of this journal are available for any citation. However, all the citations
should be clearly indicated with the title of this journal, serial number and the name of the author.

Abstracted / Indexed in:

CAS (Chemical Abstracts Service)
Database of EBSCO, Massachusetts, USA
Chinese Database of CEPS, Airiti Inc. & OCLC
Cambridge Science Abstracts (CSA)
Ulrich’s Periodicals Directory
Chinese Scientific Journals Database, VIP Corporation, Chongqing, China
Summon Serials Solutions

Subscription Information:
Price (per year):
Print $600, Online $480
Print and Online $800

David Publishing Company

9460 Telstar Ave Suite 5, EL Monte, CA 91731, USA
Tel: 1-323-9847526, Fax: 1-323-9847374
E-mail: order@davidpublishing.org

D DAVID PUBLISHING

David Publishing Company

www.davidpublishing.org
 Journal of Environmental
Science and Engineering A
Volume 1, Number 1, January 2012 (Serial Number 1)

Contents
Aquatic Environment
1 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission
Mahdi Majedi-Asl and Robabeh Jafari

Atmospheric Environment
11 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal
Station in Mexico City
Sara Rodríguez, Hortensia Reyes, Paulino Pérez and Humberto Vaquera
21 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in
Brazil
George Augusto Batista Câmara, José Célio Andrade and Paulo Sérgio de Mello Vieira Rocha

Environmental Chemistry
35 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in
Presence of Iron(II) Ions
N. Hutnik, K. Piotrowski, B. Wierzbowska and A. Matynia

Environmental Monitoring
43 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
Konstantin G. Rubinstein, Elena M. Nabokova, Roman Y. Ignatov, Maria M. Smirnova, Rafael V.
Arutunyan, Vladimir N. Semyonov, Olga S. Sorokovikova, Aleksey V. Fokin and Dmitry A. Pripachkin
62 Changes in Physicochemical Properties during Co-composting of Wastes with Different C/N
Ratios: Role of Effluents
Manel Kammoun-Rigane, Jean Charles Michel and Khaled Medhioub
68 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture in
Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
D. Albanese, M. Di Matteo and R. Pilloton
78 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
Alexandra Pavlidou
 Pollution Control
89 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site
Ágnes Dergez, Diána Bordás, Helga Kovács, Gergely Forrai, György Zsigrai, Sándor Szabó, Balázs
Magyar and István Kiss

Environment Biotechnology
99 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use
Efficiency on Corn by Radioisotope Technique
Tri Candra Setiawati
108 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products
Hamed M. El-Mashad, Ruihong Zhang and Joseph P. Greene
115 Comparison of Antioxidant Activity in Extracts of Myrtus communis L. Obtained by SFE vs.
Solvent Extraction
Paula Pereira, Maria-João Cebola and M. Gabriela Bernardo-Gil

Environmental Material
121 Feasibility of Using Sea Shells Ash as Admixtures for Concrete
Benjamin R. Etuk, Idongesit F. Etuk and Linus O. Asuquo

Environmental Art
128 Dazzling Pearl-Analysis of the Artistic Characteristics of Wu Wei’s TianTi Mountain Grotto
Jing Xue
Journal of Environmental Science and Engineering A 1 (2012) 1-10
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

The Mathematical Modeling of Self-Purification of the

Zarjoob River for Justification of Emission

Mahdi Majedi-Asl1 and Robabeh Jafari2

1. Department of Civil Engineering, University of Maragheh, Maragheh 1550477897, Iran
2. Islamic Azad University, Bonab Branch, Bonab 2142153607, Iran

Received: June 1, 2011 / Accepted: August 10, 2011 / Published: January 20, 2012.

Abstract: The main goal of this research is Rasht Zarjoob River’s self-purification study for discovering river’s water quality
reaching to water quality standards. For reaching this goal after extensive studies, a map with 1:25000 scales was prepared by
identifying all the contaminants of river water, water taking places and quality & hydrometery stations along river. Then some
samples were taken from urban, industrial and agricultural resources considering EPA standard where possible. According to these
standards and different authorities’ recommendations the least waster-existing time of river was chosen for taking samples. As well
for enhancing modeling accuracy and exact calibration, sample taking identifying factors were done three times at the end of August
half of August and middle of September. Finally, September middle data was selected for modeling. With association of
anzali-envirnoment research station, NO3, BOD, DO and temperature factors were measured along the river. At the time of sample
taking, rivers hydrogeometery factors were recognized at 6 stations and then this river calibrated with MIKE 11. For improving the
quality of Zarjoob River water to water quality standards (in this paper Pouryadgar quality standard was used) three scenario were
defined. At first scenario (depletion standard), BOD and DO parameters went over water quality standards. After studying of this
method and implementation of first scenario, the second scenario by the name of urban sewage system was analyzed in which DO
and BOD parameters still were higher than water quality standards. At the end, third scenario by the name of waste load allocation
was defined for this river in which all the BOD, DO and NO3 parameters were in the standard ranges. Ultimately, with comparison of
scenario and considering their advantages and disadvantages it was concluded that waste load allocation scenario is an appropriate
method for improving the quality of Zarjoob River quality by considering river reaction through this method and as well considering
the easiness of this method application without any need to long time and extra expenses.

Key words: Zarjoob River, quality modeling, WLA (waste load allocation), self-purification.

Nomenclature
a The adjustment coefficient of different velocities at the
cross-section
A Flow cross-sectional area (m2)
q The debit in the nutrition or uptake length unit (m2/s)
R* The hydraulic radius determined by the following
equation (-)
E Longitudinal scattering coefficient (-)
K Linear decrease rate (-)
h The Level of water with respect to a standard level (m)
Q The Cross-sectional debit (m3/s)
C The Chezy coefficient (-)
g The acceleration of gravity (m/s2)
Corresponding author: Mahdi Majedi-Asl, master, main
research fields: hydraulic of rivers and water structure
engineering. E-mail: mehdi.majedi@gmail.com.
1. Introduction
Population growth and urban and industrial
development of the country has made the importance
of paying attention to the quality of water resources
more obvious. Inability of water resources to refine
the waste materials added to them leads to many
negative effects. Besides endangering the environment
of aquatic organism, this would limit the use of the
water for urban, agricultural, sport, and recreational
purposes. For some years, urban, industrial, and
agricultural sewage has been drained into the Zarjub
River, Rasht. The river flows into the Anzali lagoon.
As a result, the health level and population of aquatic
organisms decreases every day.
2 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission

In this regard, some studies have been carried out in The equation of mass (1) and Eq. (2) are called the
Iran. For instance, Eslami [1] determined the momentum equation.
self-refinement capacity of the Kor River. By using 3
B
1 2
ETERNAL computer model, Siyahbazi [2] evaluated R * = ∫ y . db (3)
the rivers that are recipients of point and non-point Ao
contaminant qualitatively. The model employed (2) Equations of transportation, scattering, and
computer abilities to generalize the point equation for reaction
calculation of non-point contaminant using a numerical Besides the flow equations, we would have
method. Shafiee [3] modeled the manner of salinization transportation, scattering, and reaction equations, due
of the Karoon River by the tide phenomena of the to the presence of a contaminant with the
Persian Gulf using MIKE11 computer model. concentration of c(x, t). Eq. (4) is the basic equation of
In 1997, Pirasteh [4] proposed some transportation, scattering, and reaction:
recommendations for improvement of water quality of ∂c ∂c ∂ 2c
= −u + E 2 − kc (4)
the Zarjoob River in a study on the optimized ∂t ∂x ∂x
management of the Zarjoob River, Rasht. The solution of the dominant equations: The
Moreover, Khavar [5] carried out a study on the solution method used in the MIKE 11 model is the
physical and chemical evaluation of the Gouharud and limited subtraction method. By discretization of the
Siahrud Rivers of Rasht city and their effects on the dominant equations, we would have:
Anzali Lagoon. 1 ⎧1 1 n +1 ⎫
n +1
⎨ ( Q j +1 + Q j +1 ) − (Q j −1 + Q j −1 )⎬
n n
Miles [6] analyzed the sampling for portioning the
Δxxj ⎩ 2 2 ⎭
contamination load of the Iil River in Dublin, USA. (5)
AoJ + Aoj +1 1 n +1
They were willing to decrease the contamination load + (h j − h j ) = ± q
n

of the surrounding sewage purification plants such Δ2 xj Δt

n +1/ 2 n +1 / 2
that the water quality standard in terms of soluble ⎛ Q2 ⎞ ⎛ Q2 ⎞
oxygen is followed. ∂ ⎜ α
⎜ A⎟ ⎟ − ∂ ⎜⎜α ⎟⎟
Q nj+1 − Q nj ⎝ ⎠ j +1 ⎝ A ⎠ j −1
To model the Zarjoob River, we used the MIKE 11 +
model, developed by Danish Hydraulic Institute (DHI). Δt Δ2 xj
Simultaneously, WQ, AD, and HD modules were run (h nj++11 + h nj+1 ) − (h n +1
j −1 + h nj−1 ) (6)
and modeled.
+ gA 2 2
2. The Dominant Equations Δ2 xj
To evaluate the trend of contamination change trend g (θQJn +1Q nj − (θ − 1)Q njQ nj )
in the river, the following equations were used:
+ =0
C 2 AR *
(1) Hydraulic equations
Eqs. (5) and (6) are obtained by discretization of the
Water flow in the river was modeled by the
mass equation and the momentum equation,
one-dimensional Saint-Venant equations, which are
respectively.
obtained by the balance of mass (1) and momentum (2)
in a one-dimensional channel. 3. Materials and Methods
∂Q ∂A
+ = ±q (1) The inputs for modeling of the Zarjoob River were
∂x ∂t
as follows:
∂Q ∂ Q2 ∂h gQ Q (2)
+ (a ) + gA + 2 =0 y The coordinates of 115 points using maps at the
∂t ∂x A ∂x C AR *
 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission 3

scale of 1:25000; Considering vegetation and condition of the region,

y Hydraulic data: At the boundary conditions, the
debit of Behdan and Golsar stations were 0.117 and
4.251 m3/s, respectively. Moreover, the water level in
the downstream was -4 m, and the monitoring
coefficient was generally determined to be 0.025. The
cross-sectional area, debit, velocity, and height were
measured in six stations (Table 1);
y Qualitative data: The transportation coefficient in
the model was 100 m2/s, and the region’s height was
37.2 m.
The aeration and disintegration coefficients were
applied for various parameters.
The boundary conditions: At the boundary
conditions of the model, 0 and 24.515 km along the
river route were selected as the open boundary. The
water level at the Golsar station at the 24.515 km
along the river route was -4 meters. The
concentrations of primary parameters are provided in
Table 2.
(1) Hydrodynamic module
The hydrodynamic data for the six stations were
available and are presented in Table 1.
the roughness coefficient equal to 0.025 was selected
for the model. However, for calibration of the model,
the coefficient for different regions of the river was
slightly modified to make the real water level
consistent with the water level of the model.
(2) Qualitative module
The qualitative data were measured for nine stations
in the Zarjoob River route. They were the
contamination load of input contaminant resources in
point and non-point forms (Table 3).
4. Results of the First Scenario
After calibrating the hydrodynamic and qualitative
model of the Zarjoob River, the scenarios for
improvement of river’s water quality were defined.
The first scenario was defined on the basis of the
standard of sewage drainage into the route of the
Zarjoob River (Table 4). After running the scenario, it
was found that only NO3 is dependent on the
qualitative water standard. The DO and BOD of the
river were not within the permissible range of
qualitative water standards (Figs. 1-3).

Table 1 The hydraulic data of the Zarjoob River, Rasht.

Distance from Measured Cross surface Height Velocity
Station Altitude (UTM)
origin (m) charge (m3/s) area (m2) (m) (m/s)
Behdan 0 379963-4114639 0.117 0.0875 19.8 0.3
Autoban 7923 380591-4117659 1.491 3.157 9.81 0.52
Water Org. 13810 376942-4124484 2.814 3.467 3.58 0.8
Modares Bridge 18312 376489-4124789 3.114 4.380 -0.04 0.72
Zarjub Bridge 21203 375626-4127069 3.531 4.525 -2.01 0.75
Golsar 24515 374671-4128782 4.251 5.035 -4.46 0.91

Table 2 The basic concentration of parameters in the upstream of the river (Behdan station).
Parameter DO BOD NO3
Concentration (mg/L) 7.51 5 4.478

Table 3 The measured qualitative data of the Zarjoob River, Rasht.

Station Between
Water Modares Nature Keyvan Between Shariati&
Behdan Authoban Modares & Golsar
Org. Bridge defense Org. Bazarche Takhti Bridge
Parameter Iraqi Bridges
Temperature (oC) 24 25 25 25 24 23 23 23 23
DO (mg/L) 7.51 7.89 5.41 4.75 4.3 4.1 3.85 3.5 3.2
BOD (mg/L) 5 6.7 10 10 11.4 12.5 13.7 13.6 18.7
NO3 (mg/L) 4.478 3.28 1.62 1.628 1.68 1.63 1.51 1.4 1.57
4 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission

Table 4 The sewage output standards. The Iranian law 6. Results of the Third Scenario
and regulations for environmental preservation [7].
Parameter Maximum permitted level Unit In this step, the third scenario was defined by the
DO 5 (minimum) mg/L
proportioning method of contamination load to the
BOD5 5 mg/L
NH3 0.02 mg/L contaminant units along the river route using the
NO3- 45 mg/L Pouryadegar standard (Table 5). The value of BOD
5. Results of the Second Scenario parameter for the contaminant unit at the 0.4 km along
the river route was considered to be 5 mg/L. The
To determine the appropriate conditions for the Siavash water flows into the river at the 10 km along
Zarjoob River, the second scenario, i.e., using the the river route with a high BOD (130 mg/L). Thus, in
urban sewage collecting system, was defined. It order to modify the load for contaminant units at 7.5
should be noted that in this method the urban sewage km and 11 km along the river route, the BOD value
was considered to be zero. It means that as the river was determined to be 30 mg/L. Other units were
reachs the city outskirts, from the water organization allowed to drain the sewage up to BOD concentration
station on, no contamination resource was allowed to equal to 40 mg/L. The contaminant unit at the 24 km
drain sewage into the river. Thus, in this method, along the river route was allowed to drain sewage up
except for NO3, DO (only for some regions), and to BOD concentration of 40 mg/L. In this way, the
BOD, the parameters still did not consist with the BOD along the river route was consistent with the
water quality standards (Figs. 4 and 5). water quality standards (Figs. 6 and 7). As a result of

Fig. 1 DO changes along the river route (the first scenario).

 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission 5

Fig. 2 BOD changes along the river route (the first scenario).

Nitrate

Fig. 3 Nitrate concentration changes along the river route (the first scenario).
6 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission

Fig. 4 DO changes along the river route (the second scenario).

Fig. 5 BOD changes along the river route (the second scenario).
 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission 7

Table 5 A sample of the general criteria for water quality in Pouryadegar standard (the Iranian law and regulations for
environmental preservation [7]).
No. Dirty factors Discharge of surface waters (mg/L) Charge to adsorbing well (mg/L) Agricultural uses (mg/L)
1 NH4 2.5 1 -
2 BOD5 30 (on time 50) 30 (on time 50) 100
3 Solute oxygen (Min) DO 2 - 2

Fig. 6 DO changes along the river route (the third scenario).

Fig. 7 BOD changes along the river route (the third scenario).
8 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission

partitioning the contamination load, the BOD, DO,
and NO3 parameters were consistent with the water
quality standards.
7. Discussion
The results obtained by the study show that
modeling with MIKE 11 software in the Zarjoob
River was highly consistent with NO3, BOD, and DO
parameters, such that the consistency rate was up to
97% in some cases (Figs. 8-12). This is while Radwan
et al. [8] for the Molenbik River, using MIKE 11 was
is very difficult to know the quality of underground
water. This significantly affects the modeling process.
Analysis of the results of the scenarios, which were
defined for improvement of the Zarjoob River, shows
that by carrying out the scenario of partitioning the
contamination load, the water quality of the river will
be dependent on the water quality standards.
Regarding the tables, this scenario leads to desirable
results.
DO (mg/L)
10

concentration (mg/L)
shown to be 80% consistent. Moreover, in the studies 8
6
Abolghasemi-Rahimabadi [9] carried out on the DO 4
2
and BOD parameters of Zayanderud River, some
0
discrepancy was found in the re-evaluation. 0 7923 13810 18312 19153 19977 21203 23739 24515
mesurment data 7.51 7.89 5.83 5.41 6.1 5.4 3.68 3.5 3.2
Considering the results obtained from the modeling model data 7.5 7.7 5.1 4.5 4.2 4.1 3.85 3.35 3.2
for the three proposed scenarios (Table 6), the chainage (m)
following issues can be mentioned in issuing the Fig. 8 The result of calibration of DO qualitative model.
certificate for the Zarjub River:
The Zarjoob River is one of the particular rivers of NO3(mg/L)
the country, which has specific conditions due to 5
concentration (mg/L)

4
drainage of urban sewage into it as well as the high
3
level of underground water. Moreover, the 2
contamination of the river is mainly caused by urban 1
sewage, since most factories around the river have 0
0 7923 13810 18312 19153 19977 21203 23739 24515
been shut down or work temporarily due to economic mesurment data 4.478 3.284 1.619 1.63 1.783 1.921 1.355 1.623 1.527
issues. model data 4.278 3.24 1.6 1.2 1.7 1.64 1.5 1.4 1.6
In the modeling of certificate issuing, each of the chainage (m)
inputs of urban sewage is considered as a contaminant
unit. Fig. 9 The result of calibration of BOD model.

(1) The results obtained from the third scenario will

NO3(mg/L)
be considered as the basis for certificate issuing. It is
5
concentration (mg/L)

obvious that by lowering the level of sewage 4

refinement if the environmental conditions of the river 3
are preserved, the costs imposed to each production 2
1
unit will be lower. Thus, the overall cost of industrial
0
products will decrease, which is economically 0 7923 13810 18312 19153 19977 21203 23739 24515

desirable for society. mesurment data 4.478 3.284 1.619 1.63 1.783 1.921 1.355 1.623 1.527
model data 4.278 3.24 1.6 1.2 1.7 1.64 1.5 1.4 1.6
(2) Considering the particular condition of the
Zarjoob River and the high level of underground water chainage (m)
and also the frequent annual rainfall in all seasons, it Fig. 10 The result of calibration of NO3 qualitative model.
 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission 9

Model data (mg/L)

Fig. 11 Comparison between the observational inputs and the data of model for NO3 parameter.

Fig. 12 Determining the error percentage in NO3 parameter.

Table 6 Comparison of the scenarios used for improvement of the water quality of the Zarjub River, Rasht.
Parameter type Scenario 1 Scenario 2 Scenario 3 Environmental standard
Unless kilometers 0.436 and Unless kilometers 0.436 and √
DO Minimum 5
8.5√ 8.5√
Unless kilometers 0.436 to Unless kilometers 0.436 to
BOD √ Maximum 5
13.7√ 13.7√
NO3 √ √ √ Maximum 45
√ means acceptable.

Table 7 The characteristics of contaminant drainage according to the certificate.

Distance(m) Charge (m3/s) DO (mg/L) BOD (mg/L) NO3 (mg/L)
436 0.291 2 5 50
7500 0.083 2 30 50
11251 0.05 2 30 50
18200 0.05 2 40 50
24000 0.079 2 35 50
10 The Mathematical Modeling of Self-Purification of the Zarjoob River for Justification of Emission

8. Conclusions and Suggestions country’s rivers.

The studies on portioning of the contaminants load
provide the information necessary for efficient
decision-making on the necessary level of refinement
in contamination resource(s). Hence, it is suggested
that in order to achieve the water quality standards,
while issuing the certificates for contaminating units,
they are bounded to perform sampling of their sewage
output in each season, and then report the quantitative
and qualitative results to the environment organization
to make sure that they have followed the allowed
concentrations mentioned in their certificate.
Furthermore, they should be obliged to carry out
annual sampling of the output sewage. Then, the
results should be compared with the allowed
concentrations. It is hoped that by such considerations
and appropriate restrictions, we can achieve the true
feature of the Zarjub River.
Finally, it is recommended that the following
evaluations be carried out for issuing the certificate for
other rivers:
(1) Developing the standards of raw water quality
for various applications;
(2) Setting up stations for measurement of qualitative
and quantitative parameters for the country’s rivers;
(3) Identifying the contaminating units of the
In Table 7, an example of the certificate for this
study is provided.
References
[1] A. Eslami, Water qualification and self-refining capacity
of Kor River, MA Thesis, University of Shiraz, 1998.
[2] G. Siyahbazi, Modeling and consideration of distributed
dirty factors effects on rivers, MA Thesis, University of
Shiraz, 1999.
[3] Z. Shafiee, Modeling of Karoon River salting with water
up-down effect in Persian gulf, MA Thesis, Science &
Research Branch, Islamic Azad University, 2004.
[4] M.H. Pirasteh, Optimized management of Rasht Zarijub
River, MA Thesis, Science & Research Branch, Islamic
Azad University, 1997.
[5] S.L. Khavar, Physical & chemical consideration of
Goharrood and Siyahrood Rivers in Rasht City and their
effects on Talab-e-Anzali, MSc Thesis, Science &
Research Branch, Islamic Azad University, 2002.
[6] W.B. Miles, Stream Sampling for Waste Load Allocation
Application, Volume V, Center for Environmental
Research Information, 1986.
[7] The Iranian Law and Regulations for Environmental
Preservation, khorsandi, 1992.
[8] M. Radwan, P. Willems, Principle of Surface Water
Quality Modeling and Control, Hydraulics Laboratory, J.
River Management, 2003 .
[9] A. Abolghasemi-Rahimabadi, Identification and
consideration of dirty sources on quality of Zayanderood
with QUAL2E, MA Thesis, Isfahan University of
Technology, 2008.
Journal of Environmental Science and Engineering A 1 (2012) 11-20
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Selection of a Subset of Meteorological Variables for

Ozone Analysis: Case Study of Pedregal Station in
Mexico City

Sara Rodríguez1, Hortensia Reyes1, Paulino Pérez2 and Humberto Vaquera2

1. Department of Mathematics and Physics, Autonomous University of Puebla, Puebla 72570, Mexico
2. Post-Graduate Academy, Montecillo 56230, Mexico

Received: May 20, 2011 / Accepted: July 4, 2011 / Published: January 20, 2012.

Abstract: In this paper, we describe how we carried out a research using the methodology Extreme Value Theory to model a problem
of environmental pollution. Through the likelihood ratio test, we selected the best subset of environmental variables. The procedure
was applied to data of Pedregal Station, which belongs to the atmospheric monitoring system of Mexico City, managing the maximums
of each three days in the period from 1990 to 2009. The importance of these studies on ozone concentrations is that exposures for a long
period of time may cause health damage to alive/living beings, in short and long terms, mainly for people living in megacities. The
environmental variables considered in this analysis are: wind, temperature and relative humidity. The chemical variables that we took
are: carbon monoxide, nitrogen dioxide and sulfur dioxide. Although we did not examine all possible subsets, we took the most
important ones by nested subsets, comparing them using the deviance statistic. Finally, we observed that the best subsets were those
involving the time variable and, the one that contained all variables except sulfur dioxide, which lead us to think that explaining the
ozone concentration evolutions through these variables, is still a complicated case, since more accurate methodologies are discovering
new associations that are related with its creation and evolution. Environmental pollution is a very difficult problem that requires to
carry out new researches, measurements and interactions, that help to put forth the acquired knowledge into new models.

Key words: Generalized extreme value distribution, ozone, selection of variables.

1. Introduction researchers [1, 5, 6].

Ozone is the most harmful oxidant for life on earth
The Extreme Value Theory is an area of statistics
that is produced by human activities. For this reason,
dedicated to develop models and techniques that
the specialists, in conjunction with public institutions,
estimate the behavior of rare events that in some cases
have been carrying on investigations in areas related to
have catastrophic results. These belong to the
ozone and health, to measure the resistance of
long-tailed of a distribution and are far from central
individuals living in urban zones, in according with
statistics. These events exceed a threshold (maximum
interesting variable, like age orthe exposure time,
or minimum value) in a fixed time period and use the
among others [7, 8]. Several statistical methodologies
asymptotic distribution of the maximums properly
have been applied, for example multivariate models,
normalized by sequences of random variables [1-3].
non linear regressions, time series, non parametric
Extensions on this theory exist for the case of extreme
models, Bayesian and neuronal networks, copula and
modeling; one is in a continuous time [4] and the other
spatial models [9-15].
uses Bayesian methods as it has been done by some
Mexico city is a mega city since it has more than ten

Corresponding author: Sara Rodríguez, master, main
million people, who suffer of a deficiency of oxygen in
research field: statistics. E-mail: rguez.sara@gmail.com.
12 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal
Station in Mexico City
their lungs of less than 23%, in comparison with people
living in other places at sea level [16]. There are
geographical conditions in this city (a place enclosed
by mountains) that favor photochemical processes in
the atmospheric stability, causing constant thermal
inversions. Industry is located in the northern part of the
city, and winds blow from north to south, leading to
dispersion of pollutants (nitrous oxides, volatile organic
compounds, among others), and also are dragged to the
south, where complex chemical processes occur by
sunlight, which gives raise to ozone creation [17].
Mexico City’s authorities issued environmental
controls. One of them is the standard of air quality
which can not exceed 0.11 parts per million of ozone,
and that a person should not be exposed to this level for
more than an hour [18]. The Atmospheric Monitoring
System of Mexico City [19] as one of its policieshas to
verify the validity of the information and constantly
carries on analysis of its data. These information bases
are the most used in Mexico due to their reliability and
facility to obtain the data, which is available for the any
user that in turn helps diverse pollution researchers,
belonging to diverse fields on this matter like the
statistics area, or others to the environmental physical
chemistry and/or photochemical. Some of the works in
the Bayesian statistics estimate the probability that a
standard in air quality could be exceeded for a number
of times within a time interval. Assigning a non
informative initial distribution, researchers use a non
homogenous Poisson distribution with multiple
change-points and note that changes that occur in the
data are reflections of environmental actions that have
been taken by the inhabitants of Mexico City valley,
and by the Mexican government in two different years
in 2000 [20]. Another work on this Bayesian approach
is the proposal for temporary space model in the
interpolation and prediction of hourly ozone
concentrations. Temperature was taken as covariate in
some SIMAT stations, and was interpolated when data
was missing. The model incorporates spatial
covariance structure for observations and parameters
that define the harmonic components, using a dynamic
linear model. This model yields good results for short
time prognostics, but it does not incorporate transport
of pollutants, also does not respond to physical
relations and, neither involves chemical reactions that
occur at different atmospheric levels [6]. Subsequently,
the authors propose a new approach assuming that the
empirical distribution of ozone can be approximated by
a Generalized Extreme Values type distribution, whose
parameters define a time and space structure. The
location parameter has a dynamic linear model and
parameters of scale and form are depending on time. In
the analysis of Mexico City data, an example is given
in a flexible form to capture short and long term
tendencies, also depending on time. The research that
that was presented is a challenge to try variations in
time lengths or the selection of a threshold. It was
observed in the analysis that optimal selection of
thresholds depended on neighboring stations [21].
Another article that uses the classic Extreme
Generalized Values Distribution proposes a simple
model to estimate trends in time series, that fits the
ozone as the classic linear regressions models do, from
the convergence of the quantile function to a
standardized normal distribution, which allows to do
hypothesis tests and confidence interval in the
parameters of interest. The reported results confirm
that the series of data presents two change points, one
increasing before 1996 and another decreasing after
this time [22].
In the field of environmental chemistry and physics,
at the Atmosphere Science Center of the UNAM,
researchers analyze the non lineal behavior among the
precursors and ozone production in the city. They also
inquire into the causes of the rising of high levels of
ozone on a specific time. Their motivation was based
on knowing why on a weekend, that are days off,
pollution increased so much. They proposed
hypothesis, and simulated real data bases, ozone
creation through nitrous oxides, volatile organic
compounds and climatic conditions on those days. This
 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal 13
Station in Mexico City

work got to demonstrate that strong winds carried ∏

. 1 (2)
∑
pollution from Toluca metropolitan zone to the Valley
[23, 24]. The maximum likelihood estimators of the
parameters are obtained by maximizing the likelihood
2. Material and Methods given by Eq. (2), using numerical methods.
Extreme Value Theory is based on Asymptotic 3. Procedure for Selecting the Best Model
Theory, derived from Fisher and Tippet principle [25],
which establishes that there are only three types of Taking the information about observations and k
distributions which are reached on the limit of covariates, hypothesis tests are done on the covariates
maximums of stationary sequences. Fisher and Tippet coefficients to know if any of them influences the trend
found that distributions are unified in the Generalized of ozone maximums concentrations. This was achieved
Extreme Value Distribution, which has three by observing the estimator sign of the covariate
parameters μ, σ, of location, scale and shaped, coefficient, namely we need to contrast the hypothesis
: 0 : 0
respectively.
Eq. (3) is test hypothesis for covariates.
μ, σ , exp 1 1 0
(1) where represents the parameter associated with time,
μ, σ, 0 it’s associated to the trend j 1,..,k), if β 0 there
Eq. (1) is generalized extreme value distribution will be an decreasing trend, if β 0 there will be an
with µ, σ, . increasing trend and in the last case, there will be no
Reiss and Thomaset al. [1, 3] established that the tendency.
values of the shape parameter determine the The test consists in rejecting if /
distribution type. If 0 then a Fréchet distribution
type is obtained, if 0 then a Weibull distribution where is the significance level and is the i-th
type is obtained, in the case 0 a Gumbel term of the diagonal matrix | , for
distribution is obtained. The results given in Ref. [22]
allow us to use a linear regression model in order to this, variances are obtained by using the observed
analyze ozone concentrations through environmental Fisher information matrix.
and chemical variables. To accomplish this, Cox’s idea In the event that the null hypothesis is rejected with
[26] was used, who employed the regression model in a significance level α, it is evident that it exists a trend in
Generalized Pareto. In our case, we took a the observations of Y. If all covariates estimators were
covariates vector, the unknown parameter vector was zero, this would indicate that these are not influencing
expressed into exp 0. In this to explain the trend of Y.
last expression, we introduced the maximums of the Using Coles procedure of selecting nested models [1]
standardized covariates to avoid scale problems in the that supposes a model with a parametric vector
fixed time period. The likelihood function is given by: and a subset vector that is
obtained by restricting components to a value, let us
; , , ; , ,
say they take zero value. This implies to partition
vector , where the sub-vector
exp 1 exp
exp ∑ -dimensional fulfills that 0. The test involves
to obtain the estimators ( and ( to prove
the validity of the model using the relation of with
to a significance level . From where, the
14 Se
election of a Subset
S of Metteorological Variables
V for Ozone Analy
ysis: Case Sttudy of Pedre
egal
Stationn in Mexico City
C

hypothesis test is expresssed as followss: Metteorológicos (REDMET), located in diffferent placess

: : (4) on the valley. Their geoographic locations are::
Eq. (4) is test hypothessis for better model.
m nortthwest (NW) which has 177 stations, sou uthwest (SW))
The decission rule is too reject the null
n hypothessis if withh 6 stations, northeast (N NE) with 4 staations, in thee
2 wheen is the souutheast (SE) thhere are 5 stattions and in the
t center (C))
quantile (1 ) of thee distributionn , is the therre are 3 stations. We chosee Pedregal Staation which iss
difference of
o model dim mensions and . Large
L onee of the SW stations, sincce it presentss the highestt
values of D indicate thhat model explains better leveels of ozone in all the valleey, its locationn is indicatedd
data variatioons than moddel , but iff D is small, it is in Fig.
F 1.
not worth too incorporatee more variaables becausee the In
n all valleiess, thermal invversions consstantly occur,,
model has no improvem ments. The estimations with w the station to bee analyzed beelongs to Alv varo Obregónn
maximum likelihood usinng nested models lead to carry c deleegation, withh a populationn of 729,193 and 198,6477
out pairwisee testing [277, 28]. Akaikke’s informaation houusing [34]. Thhe reduction on very extreeme pollutionn
criterion (AIIC) [29]is givven by Eq. (5)): leveels has been achieved
a throuugh time [35]], with effortss
2 2log (5) formm the inhabiitants as welll as restrictio ons on dailyy
where is the t number of o parameterss estimated inn the mov vement of mootor vehicles,, accessible co osts in publicc
model M, and is the value of the tran
nsportation and a relocatiion of som me pollutingg
log-likelihoood function for
f the set with estimmated induustries, like the refineryy of Azcapo ozalco [36]..
parameters. This criterionn helps us to solve
s the probblem Unffortunately, thhe city growss by giant leap ps that causess
of choosing between rivaal models thaat are non-neested accuumulation of atmosppheric pollu utants, thiss
in terms of their
t functionnal forms. Forr a given dataa set, agggravates withh more than 18 million people thatt
several moddels can be fitted; accorrding to the AIC inhaabit in Mexico City and vehicle parking [37, 38]..
criterion thee model with smallest
s AICC is the best (m
more Theere are frequeent thermal innversions in morem than 700
negative AIC C value). perccent of the yeear, due to a nnatural pheno omenon of airr
stag
gnation, this gets
g disperseed in a graduaal way in thee
4. Applicaations
couurse of the day d and breeaks due to atmosphericc
Mexico city
c was connsidered as thhe city withh the warrming (Table 1).
hardest envvironmental problems
p in the worlda few
N
years ago, since theree were motoor vehicles that
circulated daily
d with noo restrictions at all and there
t W E

were not reggulations on industries located in the city, S

whereby theere was frequuently maximuum levels at 0.11
ppm with thhe most harm mful variable,, the ozone [30].
[
Currently, some situationns have beenn controlled withw
the collaboration of everrybody and nowadays, so it is
no longer the most pollutted in the worrld [31, 32].
The Sisteema de MonnitoreoAtmossférico (SIM MAT)
[33] of Mexxico City beggan to operaate in 1986. It is
integrated by
b the Redd Automáticaa de Monitooreo
Atmosféricoo (RAMA), thhe Red manuual de Monitoreo Fig.. 1 Geograp phic location of Pedregal station
s (PED))
Atmosféricoo (REDMA), the Redd de Depóósito which is placed in the southwestt of Mexico Citty that belongss
Atmosféricoo (REDDA), and the Redd de Parámeetros he Alvaro Obrregón delegatioon [33].
to th
 Se
election of a Subset
S of Metteorological Variables
V for Ozone Analy
ysis: Case Sttudy of Pedre
egal 155
Stationn in Mexico City
C

Table 1 S
Summarized descriptive in
nformation about
a Tab
ble 2 Names of o covariates w with its units off measurementt
Pedregal Stattion (PED) of 1990-2009,
1 n = 2073 [33]. and
d their respectivve codes, in Peedregal Stationn.
Summary of ozone
o Value Varriable code Measurem
mentunit Renamed
R
Minimum 0.029 Ozo one (O3) ppm
1st quantile 0.129 Tim
me (t) hours X1
Median 0.169 Temmperature (tmpp) DegreeC
Celcius (°C) X2
Mean 0.174 Win ndspeed (wsp) m/s X4
3rd quantile 0.215 Rellative humidity (hr) % X3
4th quantile 0.404 Carrbonmonoxide (CO) ppm X5
Maximum 2.073 Nittrogendioxide (N
NO2) ppm X6
Sullfurdioxide (SO
O2) ppm X7
We preseent a study too analyze thhe influence of 8
covariates (environment
( tal and chem mical) in ozzone
concentratioons in Pedreggal Station thhat belongs too the
metropolitann zone of Mexico
M Cityy. It shown the
measuremennt unit and thee variables coode that wereuused
(Table 2).
Using a linear regrression modeel in the scale s
parameter, for
f each grooup of 72 hoour, we tookk the
maximums of o the 7 covarriates.
Then, wee present grraphics of the t 7 covariiates
measuremennts that have different meaasuring units and
that were staandardized to avoid scale problems
p (Figg. 2).
The maxiimum likelihhood estimatoors for high and
fixed p weree obtained ussing a program m written in Ref.
[39]. Some nested modeels were estim mated for varrious
parameters , … , in
, , ̂ (6)
Eq. (6) is generalizedd extreme valuev distribuution
with covariaates.
Based on this informattion hypothessis testing caan be
done about ozone concentrations trend, in ordeer to
know the wayw some coovariates get involved inn the
model.
Fig. 3 shoows probabiliity plot, quanntile plot, dennsity
plot for the ozone conceentrations from 1990 to 2009.
Theses grapphics can be used to checck the validitty of
distributionaal assumptionns.
Table 3 shows
s the maximum
m likeelihood estim
mator
and standaard deviationn (SV) for the param meter
associated tot time. In this
t table, we
w show thatt the
estimator off time parameeter is negativve. That is to say,
ozone decreases over timme.
16 Se
election of a Subset
S of Metteorological Variables
V for Ozone Analy
ysis: Case Sttudy of Pedre
egal
Stationn in Mexico City
C

Another
A exammple with othher model is as shown inn
Tabble 4, wheree we have a negative sign in thee
2.0

coeefficients of tiime, wind speeed and nitrou us dioxide.

The
T remaininng variabless have opp posite signs,,
1.5

indiicating positive contributioon for ozone creation.

Selection
S of the Best Suubset of Vaariables: Thee
1.0

nummber of moddels to be seelected can be b big if thee

nummber of variabbles is high, in our case, there
t are 26 =
0.5

64 subsets
s to anaalyze. The moodels that will be analyzedd
are those ones nested in moddel , and thaat the biggestt
0.0

mod del contains the

t 6 covariattes, whereby we present 6
mod dels of interest (Table 5).
-0.5

Table
T 6 show ws the log-likkelihood, deviance (withh
mod del as refference) and ddegrees of freeedom for thee
testted models. The
T last colum mn of this tabble shows thee
Fig. 2 Grap phics of the 7 coovariates that were related too the resuult of testing hypothesisthhat a given model
m fits thee
ozone variablle at Pedregall station durin ng the period from
f
dataa better thann for diffferent significcance levels,,
1990 to 2009. These are O3 (dependent variable),
v tmpp, hr,
wsp, CO, SO2 and NO2 (ind dependent varriables). Due too the 0.005, 0.01 1, 0.25 and 0 0.1.
limit of spacee in the text, wee only present some graphicss. Whereby,
W it caan be observeed that the besst models

ptive analysis of distribution

Fig. 3 Decrip nal assumption
ns of GEV at Peedregal Station
n.
 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal 17
Station in Mexico City

Table 3 Model that uses ozone and time , , where MLE and SV are the maximum likelihood
estimators and their standard deviation, repectively.
Parameter MLE Standard Deviation (SV)
-2.24470 1.699 × 10-2
-0.00057 1.99 × 10-6

Table 4 Model , , where MLE are maximum

likelihood estimators, in brackets is the standard deviation with . .
Parameter MLE (Standard Deviation)
-1.9470 (1.83 × 10-2)
-0.0008 (1.99 × 10-6)
-0.1006 (6.59 × 10-2)
0.0045 (7.86 × 10-4)
-0.0514 (1.97 × 10-2)
2.6421 (9.36 × 10-1)
-1.2019 (1.26 × 10-1)
1.4989 (3.31 × 10-1)

Table 5 Nested models of interest with the covariates.

Models Covariates
t
t and tmp
t, tmp and NO2
t, tmp, NO2 and wsp
t, tmp, NO2, wsp and SO2
t, tmp, NO2, wsp, SO2 and CO

and are significant to 0.05 0.1; in the first behavior in general, since the residuals of the model
case, the model is in terms of time and in the second follow a straight line tendency, but not all graphics are
one all variables are present, except sulfur dioxide. good, for example the residual quantile plot, we can see
From Table 7 it can be concluded the models that best that the last real values were highly deviated with
fits the data are and , which agrees with the respect from the straight line. The central data of model
results obtained using the deviance criterion. shows that is a little deviated from the estimated line
Fig. 4 shows the graphics of residuals of the model (left figure), and in Model the recent data is outside
(residual probability plot). The residuals present a good of the estimated line (right figure).
Table 6 Deviance function results (D), log-likelihood, degrees of freedom (k), for Pedregal Station.
2 _1 _1
Models log-lik k
_ _
2757.335
3055.176 -595.682 7 < is better than
Rejected that is better
2095.814 1323.042 6 >
than
Rejected that is better
2351.725 811.22 5 >
than
Rejected that is better
2463.099 588.472 4 >
than
2806.472 -98.274 3 < is better than
Rejected that is better
2167.648 1775.056 1 >
than
18 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal
Station in Mexico City

Table 7 Results Akaike’s criterion, for Pedregal Station.

Models log-lik k 2 2lo g
2757.335 11 -5492.67
3055.176 4 -6102.352
2095.814 5 -4181.628
2351.725 6 -4691.45
2463.099 7 -4912.198
2806.472 8 -5596.944
2167.648 10 -4315.296

Residual Probability Plot Residual Quantile Plot (Gumbel Scale)

1.0

12
10
0.8

8
0.6

6
Empirical
Model

4
0.4

2
0.2

0
-2
0.0

0.0 0.2 0.4 0.6 0.8 1.0 -2 0 2 4 6 8

Empirical Model
Fig. 4 Residuals probability plot for and models respectively.

5. Conclusions model. In the case of comparison of the proposed

The station with the higher records of pollution is
Pedregal. Its geographical location is ideal for air flows
to carry pollutants to the southeast zone and since air
circulation is poor, due to that the valley is surrounded
by mountains, this produces higher levels of ozone
concentration, much above those health standards
allowed by Mexico City authorities. The proposed
methodology analyses the behavior of covariates
included in the model. This supposes observations on
Generalized Extreme Value distribution. It is important
to observe the value of signs in the coefficients related
with the increasing or decreasing tendency of the
nested sub-models using chemical and environmental
variables, it turns out that the models that explain better
ozone concentration are two: one of them contains time
variable and the other one contains all variables, except
sulfur dioxide.
References
[1] A. Coles, An Introduction to Statistical Modeling of
Extreme Values, 3rd ed., Springer Verlag, 2004.
[2] J. Galambos, The Asymptotic Theory of Extreme Order
Statistics, 2nd ed., Temple University Philadelphia
Pennsylvania, Krieger Publishing Company, Roberte,
1978.
[3] R. Reiss, M. Thomas, Statistical Analysis of Extreme
 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal
Station in Mexico City
Values, 2nd ed., BirkhauserVerlag, Germany,2001.
[4] S. Albeverio, V. Jentsch, H. Kantz, Extreme Events in
Nature and Society, Springer, 2005.
[5] A. Coles, E. Powell, Bayesian methods in extreme value
modeling: A review and new developments, International
Statistical Review 64 (1996) 119.
[6] G. Huerta, B. Sansó, J.R. Stroud, A spatiotemporal model
for Mexico City ozone levels, Royal Statistical Society 53
(2004) 231-248.
[7] C. Hernández, R. Téllez, A. Sanín, N. Lacasaña,
Relationship between emergency room visits
forrespiratory illness and air pollution in Ciudad Juarez,
Chihuahua, Public Health in Mexico 42 (2000) 288-297.
(in Spanish)
[8] Population Reference Bureau (PRB) [Online], Nexos,
2002,
http://www.prenatal.tv/lecturas/efectos%20del%20medio
%20ambiente%20en%20la%20salud%20infantil.pdf.(in
Spanish)
[9] X. Guyón, Modelación para la Estadística Espacial,
Revista de Investigación Operacional 31 (1) (2010) 1-33.
(in Spanish)
[10] A. Davison, Spatial Extremes and Applications:
Discussion Outcomes [Online], 2009,
http://extremes.epfl.ch/files/content/sites/extremes/files/u
sers/111184/public/WorkshopJuly09/ExtremesDiscussion
sOutcomes.pdf.
[11] S. Porter, T. Rao, G. Zurbenko, M. Dunker, T.G. Wolff,
Ozone air quality over north America: Part IIAn analysis
of trend detection and attribution techniques, Air and
Waste Management Association 51 (2001) 283-306.
[12] A. Quintela del Río, M.F. Fernández, Extreme values
analysis of ozone concentrations using nonparametric
methods, The Science of the Total Environment 409 (2010)
1123-1133.
[13] H. Sang, A. Gelfand, Continuous spatial process models
for extreme values, Journal of Agricultural, Biological,
and Environmental Statistics 15 (2009) 49-65.
[14] M.L. Thompson, R. Joel, H.C. Lawrencw, G. Peter, D.S.
Paul, A review of statistical methods for the
meteorological adjustment of tropospheric ozone,
Atmospheric Environment 35 (2001) 617-630.
[15] T. Wolff, M. Dunker, T. Rao, S. Porter, G.I. Zurbenko,
Ozone air quality over north America: Part I-A review of
reported trends, Air and Waste Management Association
51 (2001) 273-282.
[16] J. Molina, L. Molina, The impacts of magacities on air
pollution, in: Environmetal Aspects of Urbanization
Conference, Goteborg, Sweden, 2004.
[17] A. Young, E. Betterton, L. Saldivar de Rueda, Atmósfera,
México, UNAM 10 (1997) 161-178.
[18] NOM, Modification Mexican Official Standard
19
NOM-020-SSA1-1993, Official Journal of the
Federation-Mexico, 2002.
[19] SIMAT (Sistema de Monitoreo Atmosférico de la Ciudad
de México), Agricultura No. 21, Primer Piso, Col.
Escandón, Del. Miguel Hidalgo, C.P. 11800 [Online],
2010, http : //www:sma.df.gob.mx./simat2/.
[20] J. Achcar, E. Rodríguez, G. Tzintzun, Using
non-homogenous Poisson models with multiple
change-points to estimate the number of ozone
exceedancesin Mexico City, Environmetrics 22 (1) (2009)
1-12.
[21] G. Huerta, B. Sansó, Time-varying models for extreme
values, Environmental and Ecological Statistics 14 (2005)
285-299.
[22] H. Reyes, H. Vaquera, J. Vilaseñor, Estimation of trends
in high urban ozone levels using the quantiles of (GEV),
Environmetrics 21 (2010) 470-481.
[23] A. García, A. Jazcilevich, L. Ruíz, R. Torres, L. Suárez, N.
Reséndiz, Ozone weekend effect analysis in México city,
Atmósfera 22 (3) (2009) 281-297.
[24] R. Muñoz-Cruz, M. Granados-Gutiérrez, M. del C.Jaimes
Palomera, Análisis del comportamiento semanal del ozono
en la Zona Metropolitana del Valle de México en el
periodo 1990-2006 [Online], 2007,
http://www.sma.df.gob.mx/simat/informes.tec.htm
[25] J. Pickands, Statistical inference using extreme order
statistics, The Annals of Statistics 3 (1) (1975) 119-131.
[26] W. Cox, S. Chu, Assessment of interannual ozone
variation in urban areas from a climatological perspective,
Atmospheric Environment 30 (1996) 2615-2625.
[27] G. Casella, R. Berger, Statistical Inference, 2nd ed.,
Thomson Learning, Duxbury, 2002.
[28] V. Hogg, W. McKean, T.A.Craig, Introduction to
Mathematical Statistics, 6th ed., Prentice Hall, Jersey,
2007.
[29] H. Akaike, A new look at the statistical model
identification, IEEE Transactions on Automatic Control
19 (6) (1974) 716-723.
[30] Mexican Official Regulation, NOM-020-SSA1-1993
[Online],
http://www.sma.df.gob.mx/sma/links/download/archivos/
nom-020-ssa1-1993.pdf.
[31] Blacksmith Instituto, Los lugares más contaminados del
mundo, Blacksmith Instituto Web site, 2010,
http://blog.nuestroclima.com/?p=183/2010/12/31/los-luga
res-mascontaminados-del-mundo/.(in Spanish)
[32] H. Bravo, La contaminación atmosférica por ozono en la
zona metropolitana de la Ciudad de México: evolución
histórica y perspectivas, in: IX Comisión Nacional de los
Derechos Humanos, 1992. (in Spanish)
[33] Norma OficialMexicana NOM-020-SSA1-1993, 2004,
SIMAT [Online],
20 Selection of a Subset of Meteorological Variables for Ozone Analysis: Case Study of Pedregal
Station in Mexico City
http://www.sma.df.gob.mx/simat/proteccion/(ante)nom-0 http://projects.wri.org/sd-pams-database/mexico/program
20-ssa1-1993.pdf. (in Spanish) -improve-air-quality-mexico-city-metropolitan-area-proai
[34] National Institute of Statistics and Geography (INEGI) re
Website [Online], http://www.inegi.org.mx/default.aspx?. [37] A.D. Jazcilevich, R.G. Agustin, R.S. Gerardo, A study of
[35] S. Rodríguez, H. Reyes, H. Vaquera, G. Linares, air flow patterns affecting pollutant concentrations in the
Modelación estadística en contaminación del aire para dos central region of Mexico, Atmospheric Environment 37
estaciones vecinas del estado de México, Aportaciones y (2003) 183-193.
Aplicaciones de la Probabilidad y la Estadística, [38] Censo, Instituto Nacional de Estadística y Geografia
Benemérita Universidad Autónoma de Puebla Dirección Website, 2010,
General de Fomento Editorial, Primera Edición 3 (2009) www.inegi.org.mx/Sistemas/temasV2/Default.aspx?s=est
43-53. (in Spanish) &c=17484.
[36] R.A. Montiel, O.A.M. López, V. Lichtinger, M.J. Frenk, [39] Stephenson y Gillel, ismev Package: Extreme Values in
Program to improve air quality in the Metropolitan area R", R Foundation for Statistical Computing [Online], 2004,
Mexico 2002-2010, World Resources Institute [Online], http://www.r-project.org.
Journal of Environmental Science and Engineering A 1 (2012) 21-34
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Carbon Capture and Storage Technology: World

Overview and Potential to Mitigate GHG in Brazil

George Augusto Batista Câmara1, José Célio Andrade1 and Paulo Sérgio de Mello Vieira Rocha2
1. Federal University of Bahia, Salvador CEP 40.210-630, Brazil
2. University Salvador (UNIFACS), Laureate International Universities, Salvador CEP 41.950-275, Brazil

Received: July 2, 2011 / Accepted: August 10, 2011 / Published: January 20, 2012.

Abstract: The need for global actions to combat global warming resulting from anthropogenic emissions of greenhouse gases (GHG)
has moved both the public and private sector organizations. Capture and storage of carbon dioxide (CO2) in geological reservoirs
(CCS) has been proposed as a major GHG mitigation technology in the short and medium terms. However, the current forms of
incentives for the use of CCS are related to the private sector (oil and coal industry) and funds. One way to encourage the use of CCS
on a large scale is the carbon market such as the initiative in the voluntary market. While the regulated market discusses about the
inclusion of CCS as a clean development mechanism (CDM) which should be completed for post-Kyoto period. The purpose of this
paper is to identify the international status and global perspectives of using CCS technology in order to verify its potential to mitigate
GHG in Brazil. The methodology is composed of exploratory, descriptive and analytical research which focuses on qualitative and
quantitative strategies and techniques. This paper contains an introduction about global environmental governance (GEG) climate
and the importance of CCS to the world. The literature about the CCS was reviewed and identified the key players as well as current
forms of incentives for the use of CCS on a large scale in the world. As results are presented, this article will point out the position of
the Brazilian Government about CCS and the inclusion of CCS as CDM, the potential use of CCS in Brazil and the identification of
the CCS key players in Brazil. Finally, the paper presents the Brazilian opportunities for using CCS on a large scale and the scenario
in which Brazil would be supportive of the large scale use of CCS.

Key words: Capture and storage of carbon dioxide (CO2) in geological reservoirs, CCS technology, GHG, carbon market, Brazil.

1. Introduction several steps have been taken towards the

establishment of a governance system of production,
Scientific analysis on the effect of anthropogenic
consumption and environment more balanced.
emissions of greenhouse gases (GHG) emissions and
However, even with all the dedicated effort, the
its consequences on climate change gained public
reflections caused by human action are latent in the
notoriety in recent decades. Initially, the attention
environment and increasingly worrying.
given to environmental issues was focused on issues
Due to the relationship between
involving the use of available natural resources and
production/consumption/environment the private
pollution control, but always placed in this context the
sector is a major interest in issues relating to global
economy.
environmental governance (GEG), since the
The relationship between environment and
establishment of legal framework and legal
economics has been a subject discussed since the first
infrastructure has a direct impact on company
United Nations Conference on the Human
guidelines and their production compromising their
Environment in Stockholm in 1972. After Stockholm
competitiveness. According to Porter and Brown [1]

Corresponding author: George Augusto Batista Câmara, the productive sector has historically been seen as an
master, Ph.D. candidate, main research fields: capture and
storage of carbon dioxide (CO2) in geological reservoirs and oil
opponent of national environmental policies and
& gas regulation. E-mail: george@camaraconsultoria.com.br.
22 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
global and environmental issues as a threat to
competitiveness due to the imposition of additional
costs. The role of corporate actors in GEG
traditionally took place in an indirect way, using its
political influence with state actors to veto or weaken
environmental schemes, with shares of lobbying.
An action in partnership between the public and the
private sector which directly influenced the climate of
GEG was the implantation of the Clean Air Act’s
Acid Rain by U.S. Government to combat acid rain.
This program was the main reference to the creation
of carbon credits market existing established under the
Kyoto Protocol.
In a broader analysis of the current status of policies
regarding climate change, Helm [2] stated that
although scientific evidence of climate change
increase, the political responses have been so far little
or no impact on the accumulation of emissions. As far
as technological questions is concerned Pacala and
Socolow [3] argued that humanity already possesses
scientific, technical and industrial know-how to solve
the problem of carbon and climate by the second half
of the 21st century. They show the seven (07) wedges
for climate change mitigation. The wedges are divided
into two main groups: the first, “reduction of CO2
emissions free or processes of CO2”, containing the
following technology solutions: energy efficiency,
renewable energies, decarbonization of fossil fuels
(coal/oil/gas), hydrogen and nuclear energy. The
second group, with the purpose of “capture and store
CO2” emitted in industrial processes or in the
atmosphere, composed of CO2 capture and storage and
conservation of forests.
In the technological options outlined by Pacala and
Socolow [3], capturing and storing CO2 is the
geological storage of CO2. This technology is
considered critical to reducing global emissions and
has been widely researched and developed by several
countries for the purpose of its use in large scale. It is
assumed that the domain of this technology will result
in competitive advantages for countries, including
Brazil, due to the current global environmental
governance climate established.
This article presents a world overview of the
geological storage of CO2 in order to evaluate the
potential of this technology to mitigate GHG in Brazil.
To achieve this objective, its methodology is
composed of exploratory, descriptive and analytical
research which focuses on qualitative and quantitative
strategies and techniques. For the collection and
analysis of secondary data, it was used the traditional
resources for research. Initially, it was reviewed the
literature about the technology of geological storage of
carbon dioxide. In addition, it was analyzed
institutional documents such as reports, studies and
projects. Primary data were collected through
consultations with specialists, such as academic
researchers, researchers, researchers from companies
and stakeholders from the Governments, NGOs and the
private sector, through interviews; indirect observation
of the reality studied, aiming at building space research
cooperation, participation in the agenda of discussions
and lectures on the subject of study, participation
forums with experts on the subject and direct
observation, through participation, with the seal of the
actors involved in official meetings and subsequent
synthesis of the results achieved.
2. Literature Review
2.1 The Storage of CO2 in Geological Reservoirs
According to Marchetti et al. [4], the idea of
capturing CO2 from fossil fuel combustion and storing
it in the ocean or underground had already been
floated in the late 1970s but it was only in the
mid-1980s, as worries about climate change spread
more widely in the scientific community and reached
the attention of international policy makers, that work
on CCS really began. From the early the International
Energy Agency (IEA) played an increasingly
important role in stimulating research around CCS.
Throughout the 1990s active international networks
from industry, academia and government pushed
 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
forward understanding of CCS. Fossil fuel producers
and governments were major sponsors of these
investigations, which focused on technologies for
capture and transport, assessments of storage potential,
and the modelling of costs [4].
The importance of large-scale use of technology
CGCS to reduce CO2 emissions is an issue throughout
the international community. As studies of the
International Energy Agency [5] perspectives and
technological scenarios capture and storage of CO2
can control the emissions in the short to medium term,
the study says that among the current technologies
available or likely to become commercially available
within two decades the use of CO2 capture and storage
in the industrial, fuel transformation and power
generation sectors accounts for 20% to 28% of the
CO2 savings, except in the No CCS scenario, based on
a horizon to 2050. CCS at coal-to-liquids and
gas-to-liquids plants, at refineries, and in the
production of hydrogen contributes between 3% and
5% of the CO2 emissions reduction depending on the
scenario, while CCS in industry contributes 4% to 6%.
The greatest potential for the application of low-cost
CCS is in power generation, where CCS contributes
between 12% and 18% of the CO2 emission
reductions depending on the scenario. These situations
are showed in Fig. 1 and Table 1.
M CO2
60,000
50,000
40,000
30,000
20,000
10,000
Fig. 1 Global CO2 emissions in the baseline scenario, ACT scenarios and TECH plus scenario. Index by Ref. [5].
23
The G8 (group of the eight major industrial powers
in the world) decided in June 2008 that would support
the recommendations of the IEA and the Leadership
Forum on Carbon Sequestration (CSFL) for execution
of 20 projects of CGS on large scale, because the G8
believes that the CGS has a critical role in combating
climate change and energy security challenges.
The Intergovernmental Panel on Climate Change
(IPCC) [6] defines the capture and storage of CO2 in
geological reservoirs as a process that consists in the
separation of CO2 emitted by stationary sources,
transportation and long-term storage in geological
reservoirs, isolating it from the atmosphere. It is
possible to separate the CO2 emitted from stationary
sources, process it to its liquid form and transport it by
pipeline, road or by sea to geological reservoirs such
as saline aquifers, oil or gas fields, among others.
Much has been made to develop technologies for
separating CO2 from gaseous streams emitted by
stationary sources, being the most costly phase and
increased energy consumption is to capture the CO2.
The CO2 capture technologies can be divided into:
post-combustion, pre-combustion and combustion
with oxygen. Regarding the employment of
technology to capture CO2, it exists: absorption,
adsorption, semipermeable membrane separation,
cryogenic separation and chemical looping [7].
24 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
Table 1 CO2 emissions by region in the baseline, map and no CCS scenarios, 2003 and 2050. Index by Ref. [5].
Baseline scenario
2003 2050 Map
(MtCO3)
OECD 12969 21949 8772
Transition economies 2534 3953 2283
Developing countries 9020 32120 14913
World 24523 58022 25968
With the use of CO2 capture technologies can
prevent the emission of this gas in the atmosphere.
But one of the main points at issue is what to do with
CO2 after his capture. Some technological solutions
have emerged as the destination of choice for CO2,
among them, storage in geological reservoirs.
The injection of CO2 in geological reservoirs is a
process that has been used in some industrial sectors.
There are technologies in the oil industry for enhanced
recovery of oil or gas (EOR/EGR) using some
chemicals to the injection, among them CO2.
According to the IPCC the main options for
geological storage of CO2 are: depleted oil and gas
reservoirs, use of CO2 in enhanced oil recovery, deep
unused saline water-saturated reservoir rocks, deep
unmineable coal seams, use of CO2 in enhanced coal
bed methane recovery and other suggested options
(basalts, oil shales, cavities) [6].
The term covers CCS (carbon capture storage) from
capture and transport of CO2 to its geological storage.
Since the term CGS (carbon geological storage) is
specific to the injection and storage of CO2 in
geological reservoirs. Additionally, some Brazilian
authors combine the terms CCS and CGS and use the
term CCGS (carbon dioxide capture and geological
storage) to portray the capture of CO2, the transport of
CO2 and its storage in geological reservoirs.
2.2 CCS International Players
Several players are involved in the domain of CCS
technology, both internationally and nationally,
especially governments, private companies,
non-governmental organizations (NGO) and research
ACT scenarios
Map No CCS
No CCS
Reduction below baseline scenario in 2050
(%) (%)
9943 -60 -55
2734 -42 -31
17021 -54 -47
29689 -55 -49
centers. Each player has a specific interest in
technology.
There are some indicators that reflect the interests
of the actors in CCS, for example in relation to the
private sector, the level of interest in each company
can be found in relation to their investments. For the
research, it can be seen that the attention given to
technology with the finding of some area created
specifically for research centers to the CCS or the
number of researchers. For NGOs, they can check
your level of interest involving the main forums for
discussing technology and its position also with the
amount of publications about CCS. Finally the
Government, the level of interest of a government in
CCS can be verified by their structural actions such as
developing regulatory framework for the use of CCS,
the number of projects underway or under preparation
in the country, implementation of tax or rates, the
creation of funds, the amount of resources devoted to
research on CCS and active participation in specific
forums about the CCS.
It stands to reason that some countries are more
advanced stage in the domain of CCS the United
States, Canada and Norway. Considering that, they
have ongoing projects, providing funds, encouraged
research, established a regulatory framework and
actively participate in Forums about the CCS.
In the United States many projects of geological
storage of CO2 are linked to EOR projects. To get a
better sense of the amount of existing projects in the
United States, Fig. 2 shows their locations. It is
notable that Fig. 2 shows the locations of sources of
industrial emissions and natural sources of CO2 that
 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
Fig. 2 Location of projects underway in the U.S. with CO2 EOR and pipelines. Index by Ref. [8].
serve as input to the EOR. The entire network of
pipelines are built to attend to the EOR projects and
also natural sources of CO2 which has not yet been
explored.
Injection technology products in geological
reservoirs are not recent. The activity of injected
substances and waste in the underground of the United
States is held for more than 50 years. This practice is
seen as essential to many human activities, such as oil
production, chemical, food, industrial and mineral
production. According to the FY 2000 National
Injection Well (from 11/07/2006)1 there are currently
more than 400,000 wells with substances injected into
the country [9].
Due to its experience in underground injection of
substances, the United States has a technological and
regulatory structure that favors the development of
CCS projects. It is notable that this experience not
only refers to the injection of CO2 but also other
substances, however, the context favors the domain of
CCS, especially with the experience with EOR in oil
1
Information collected on the EPA website
http://water.epa.gov/type/groundwater/uic/upload/ 007_12_12_
uic_UIC-Inventory.pdf, in 10/15/2010.
25
and gas industry.
It is noteworthy that in the initiative of some
provinces in Canada and states in the United States,
the Weyburn Project started in 1998. The project is
located in the Canadian state of Saskatchewan. The
project uses CO2 captured from a power plant and
injected into mature oil fields. Even the initial
objective is to recover the oil and gas, after which the
injection of CO2 will continue for the purpose of
storage. A task force was established to handle
specific regulatory issues to issues related to CCS on
interstate oil and gas compact commission (IOGCC)
[7] and implementation of the project is done by the
University of Regina [8].
In the European Union there are ongoing projects,
with emphasis on the Sleipner Project located in
Norway. This project aims to store CO2 in deep saline
aquifer located in the North Sea and has been in
operation since 1996. Apart from Sleipner, other
projects in Europe should also be highlighted, such as
Kalundborg in Denmark, Mid-Norway also in Norway,
Schwarze Pumpe in Germany and the Valleys of the
United Kingdom. Another important project is that of
in Salah in Algeria, one of the few located in an
26 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
undeveloped country.
Among the European initiatives have highlighted
the project NearCO2, that is investigating European
public perceptions of CCS via the case studies (both
CCS and analogous energy infrastructure), surveys,
and focus groups. The NearCO2 project is funded
under the FP7 Programme of DG Research at the
European Commission. The overall aim of the project
is to investigate communication strategies that are
designed to convey to stakeholders and to the public
advantages and risks of CO2 capture and storage. These
same strategies may also be used to involve these
parties in local decision-making on CCS projects [10].
The world has been structured for large scale use of
CCS. Players of the private sector, governmental and
non-governmental are seeking to deepen the field of
CCS technologies, as well as issues relating to funding
and sustainability. Among the existing institutions, it
can be highlighted the CDP (Carbon Disclosure
Project-relating to governance with private
companies), the UNFCCC (UN Framework
Convention on Climate Change), IPCC
(Intergovernmental Panel on Climate Change), IEA
(International Energy Agency) and UNDP (United
Nations Development Programme). Specifically
referring to technologies of CCS: Cooperative
Research Centre for Greenhouse Gas (CO2CRC),
Carbon Sequestration Leadership Forum (CSLF),
Bellona Foundation, Carbon Capture & Storage
Association (CCSA), Carbon Capture and
Sequestration Technologies Program at MIT, The
Centre for Innovation in Carbon Capture and Storage
(CICCS), Scottish Centre for Carbon Storage,
University of Regina (International Performance
Assessment Centre for the Geologic Storage of CO2,
IPAC-CO2). As for the business sector include:
Gassnova, SINTEF, Global CCS Institute,
Schlumberger, GEOGREEN, DNV and BIG CO2.
Specifically, the oil industry: Shell, Staloil,
StatoilHydro, Chevron, ExxonMobil, Encanada, BP,
Total and Petrobras.
2.3 Forms of Incentives for the Use of CCS on Large
Scale
The CCS has its potential to mitigate GHG, which
is globally recognized. Just as the economic impacts
are arising from the use of CCS and possible increase
of prices of products related to the CCS, for example
energy generation.
With the perception that the environmental impacts
generated by production processes can be minimized
through economic instruments, the creation of
economic instruments and their use in order to
encourage the use in large scale of various
technologies with lower emissions GEE was an
advance. One of the main instruments created is the
market for carbon credits.
Currently, there are two markets in which carbon
credits are traded. They are regulated carbon credit,
created from the Kyoto Protocol, and the market for
voluntary carbon credit. Besides the market of carbon
credits other form of incentive used to fund CCS
projects are the funds, the example of Norway which
receives funds through taxation on CO2 emissions.
Below are shown these three forms of international
incentives for large-scale use of CCS technologies and
how these forms of incentives are being used or not
for the diffusion of CCS.
2.3.1 Regulated Market
Since the 1990s, the world is mobilized on the
issues of climate change. In 1992, in Rio de
Janeiro-Brazil has signed the United Nations
framework convention on climate change (UNFCCC),
and in 1997 was ratified the Kyoto Protocol which
established which is called today the climate change
regime.
The Kyoto Protocol establishes the mechanisms for
flexibility regarding the emissions of greenhouse
gases, they are: the joint implementation and
emissions trading [11]. These mechanisms of the
Kyoto Protocol have fueled the establishment of the
Carbon Credit Market Regulated, especially after its
ratification in February 2005. The regulated market
 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
(called compliance) has all its established rules and
frequently reviewed and discussed at regular meetings
and with the participation of several players.
The carbon credit market created by the Kyoto
Protocol is the most representative today. The value
per ton of CO2 traded in the regulated market is
greater than a ton traded in voluntary markets. The
CCS is not considered as a technology of the clean
development mechanism (CDM), therefore, it is not
possible to obtain carbon credits in the regulated
market with the CCS for that to happen it is necessary
that the CCS will be adopted in the UNFCCC.
2.3.2 Voluntary Market
Another action that is being made in relation to the
carbon credit market is the creation of the voluntary
markets. These markets are formed by groups or
sectors that have no obligations as to reduce their
emissions or by companies located in countries not
signatories to the Kyoto Protocol as the United States.
One example is the Chicago Climate
Exchange-CCX ®. The CCX ® has attracted many
large companies, but this initiative shows signs of
exhaustion because United States does not have an
emissions reduction goal established. Some regional
initiatives in the United States gained strength due to
the possible establishment of state targets such as
American Regional Gas Greenhoue Initiative (RGGI)
and the Western Climate Initiative (WCI) gained force
due to possible establishment of state targets for
reducing emissions.
Obtaining carbon credits in the regulated market or
the voluntary market follow a classification by sector.
These sectors are studied and selected to be developed
methodologies/standards that will be capable of
obtaining credit.
In the regulated market sectors selected are: the
energy industry (renewable/non renewable), energy
distribution, energy demand, manufacturing industry,
industrial chemical, construction, transportation,
mining/mineral production, metals production,
fugitive fuels (oil, gas and solid), fugitive emissions
27
from production and consumption of halocarbons and
sulfur hexafluoride, solvent use, waste treatment and
disposal, afforestation and reforestation, and
agriculture. Altogether there are fifteen (15) areas that
harbor three thousand three hundred seventy-seven
(3377) 2 projects in various UNFCCC approved
methodologies [12].
In the voluntary market each institution selects the
sector that will work and if necessary develop
standards or use the methodologies of the UNFCCC.
However, there is the ANSI-GHG-PR-706 that orients
the definition of sectoral scopes developed by
American National Standarts Institute [13].
Both the sectors selected by the UNFCCC as
suggested by ANSI GHG-PR-706 are guidelines for
the voluntary market institutions because such
institutions have the freedom to act in different scopes.
Thus, market institutions can voluntarily choose the
area of operation that can be determined by market
demand.
For example the Voluntary Carbon Standard (VCS)
follows the structure of groups and scopes of projects
established in ANSI, develops its standards and
accepts the methodologies approved under the
UNFCCC (CDM and JIs) and developed by the
Climate Action Reserve.
The Gold Standard (GS) has as reference the scopes
defined in the UNFCCC and adopts the approved
CDM methodologies, and develops their own
methodologies. Another, which also adopts the
approved methodologies under the CDM, is the
American Carbon Registry (ACR) which also
develops its own methodologies for non-CDM scopes
such as reduced emissions from deforestation and
degradation (REDD) and CCS.
The ACR has certified carbon credits from CCS
projects. Three certified projects were all located in
the U.S. and made to oil companies in applying for the
enhanced oil recovery (EOR) with capacity to mitigate
2
Information collected on the UNFCCC website:
http://cdm.unfccc.int/Statistics/Registration/RegisteredProjByS
copePieChart.html, in 02/ 25/2011.
28 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
5.6 Mt CO2 it was implemented between 2004 to 2008.
The VCS is also been demonstrated to certify
interested CCS projects, however, there wasn’t still
demand and in consequence did not develop
methodologies for CCS.
According to the report the WorleyParsons Services
Pty Ltd produced for the Global CCS Institute, in
2009 there were 213 active or planned the projects,
which were examined by further queries and their
scale CCS project type. Of the 213 active or planned
projects, 101 are of commercial scale. Of these 62
projects were considered the integrated, which
demonstrate the entire CCS process chain of CO2
capture, transport and storage. Seven of these are
already in operation [14]. Actually there are around 13
CCS projects worldwide have the potential to mitigate
CO2 of 135.26 Mt by 2050. Fig. 3 shows the location
of projects in the world.
The principal source of resources for the use of
CCS technologies is the funds. Existing funds to
finance projects for CCS include: American
Governments USA Recovery and Reinvestment Act
(ARRA), the Canadian Government’s Action Plan,
Fig. 3 Geologic storage and related projects in operation or proposed around the world to address climate change. Index by
Ref. [15].
European Economic Recovery Plan (EERP) and
Australian Coal Association’s COAL21 Fund. Much
of the resource destined for projects of CCS are the
governments of developed countries like the United
States, Canada, Japan, Australia and European
countries with emphasis on Norway. It is estimated to
be available from approximately $17 to $20 billion of
resources for projects of CCS [14].
3. Results and Discussion
Since 2007, some emerging countries published
their compromises to reduce GHG emissions. The first
was China in 2007 (National Climate Change
Program), then in July 2008 were India (National
Action Plan on Climate Change-NAPCC) and South
Africa (Long Term Mitigation Scenarios-LTMS), in
December 2008 was Brazil (National Plan on Climate
Change-PNMC) and the last country that has
published its targets was Mexico.
Among the emerging countries cited, only South
Africa inserted the CCS in its GHG reduction program.
It can be seen more interest from developed countries
by the CCS. A bigger incentive for the emerging
 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
countries could be the inclusion of CCS as CDM. The
UNFCC is the main regulatory ambient in which
countries and international players are asked about
technologies of GHG mitigation
The possibility of a CGS project being included as a
CDM was discussed during the first session of the
11th Conference of the Parties to the United Nations
Framework Convention on Climate Change (COP 11)
and during the first Meeting of the Parties to the
Kyoto Protocol (MOP 1), in Montreal, Canada in
December 2005 [16]. Since then, the CCS has been a
theme of discussion in all subsequent conferences.
In Copenhagen in December 2009 (COP 15/MOP 5)
the importance of CCS as a reduction technology was
officially recognized [12].
Besides the meetings, the UNFCCC asked the
opinion of some governments and players regarding
the inclusion of CCS as CDM. Of the Governments
consulted by the UNFCCC only Portugal in 2007 and
Fig. 4 Position of countries and other players and the inclusion of CCS as CDM.
29
Brazil in 2008 and 2010 were against the inclusion of
CCS as CDM. For the other players only Greenpeace
was against the CCS. There is a larger amount of
players than the Governments which are against the
inclusion of CCS as CDM. The Fig. 4 summarizes the
position of Governments and players consulted by the
UNFCCC until 2010 on the inclusion of CCS as a
CDM.
The fact that the UNFCCC is a forum in which the
participants have a voice in reflects the position of the
key players about the CCS, and its inclusion as CDM.
Fig. 4 confirms what was said by Meadowcroft and
Langhelle [4], just as Bellona has become a
determined advocate of CCS, so Greenpeace
International emerged as one of its most consistent
critics. In addition to the positioning of Bellona and
Greenpeace it is also worth mentioning the emphasis
in favor of inclusion of CCS as CDM by CSFL and
the Government of Saudi Arabia.
30 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
In December 2010 at COP 16 (Cancun/Mexico)
CCS was recognized as an important technology to
mitigate GHG emissions and it was considered
eligible under the CDM’s. However, some steps still
must be taken to validate the inclusion of CCS as
CDM for post-Kyoto period.
These past February and March, Indonesia, Qatar
and Japan favored for the inclusion of CCS as CDM
despite the fact that the Greenpeace has opposed to it.
3.1 Brazilian Government’s Position on the Inclusion
of CCS as CDM
The Brazilian Government is not against the use of
CCS, however, believes that CCS is not eligible as
CDM for several reasons, such as the lack expertise in
the implementation of CCS in developing countries,
the high costs of dissemination and technology
transfer, the evaluation of environmental impact and
the process is capital intensive and technology [12].
Due to the format of the CDM, the CCS has
features that are not compatible as leakage, project
boundary, responsibility and long term stay. Other
issues highlighted are the potential economic impacts
and market conditions have not been evaluated [12].
The Brazilian government believes that CDM
incentives should be utilized to promote cleaner
technologies and renewable for not promoting the use
of oil, gas and coal. The incentive for the use of CCS
in developing countries could take place through
mechanisms specific financial, funding and
partnerships under the UNFCCC and not as a
compensation mechanism. The inclusion of CCS as
CDM can reduce the search for other more appropriate
financial mechanisms under the UNFCCC or
government policies [12].
The risks of CCS would only be minimized with the
consideration of possible commercial application of
CCS. The only mature market for the use of CCS
identified in the IPCC report is the oil industry for the
implementation of enhanced oil recovery (EOR). The
Government of Brazil believes that the CDM was not
designed to subsidize the production of oil and natural
gas. This type of project can not depend on or be
complemented by incentives from the CDM. The
producers of fossil fuels do not need this kind of
subsidy to the oil price above U.S. $80/bbl. In
addition, these companies have expertise and
significant investments in the area of CCS. The CDM
should not be intended to encourage the extraction of
methane in coal mines [12].
The CCS is typically a transition technology to be
used in the passage of an economy based on fossil
fuels to a low-carbon intensity. The Government of
Brazil recognizes that the CCS may be useful for a
long time. Thus, the CCS technology could be a
bridge until countries can have full confidence in
renewable energy, but the CCS under the CDM would
result in perverse incentives for increased production
of fossil energy in developing countries which would
enhance the existing technological gap between the
developed and developing world [12].
3.2 Potential of CCS to Mitigate GHG in Brazil
After the COP 15 the Brazilian Government ratified
the Law of the National Policy on Climate Change.
According to this law, “it will adopt, as a national
commitment voluntary actions to mitigate emissions
of greenhouse gases, in order to reduce from 36.1%
(thirty-six integers and one tenth percent) and 38.9%
(thirty-eight whole-nine percent) projected emissions
by 2020” [17].
Even Brazil with production-based energy
generation by hydroelectric plants in 2007 was
published the National Energy Plan 2030 [18] that
shows the need to expand power generation by power
plants based on coal, nuclear, biomass and natural gas
between the years 2005 to 2030 and increased
participation of:
 Coal—an increase of 1.3% (from 1.7% in 2005
to 3% in 2030);
 Nuclear—an increase of 2.2% (from 2.7% in
2005 to 4.9% in 2030);
 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
 Biomass—an increase of 3.2% (from 0% in 2005
to 3.2% in 2030);
 Natural gas—an increase of 4.9% (from 3.8% in
2005 to 8.7% in 2030).
With the increase in power generation by power
plants, according to the National Energy Plan 2030,
GHG emissions in the next 25 years should have a
range of more than 770 Mt/CO2 in 2030, with annual
growth rate of 3.5% over 2005 [18].
Brazilian potential of CO2 storage in geological
reservoirs is approximately 2,000,000 Mt [19]. The
use of CO2 capture and storage in the fuel
burn-industry, industrial processes and fugitive
emissions sectors accounts for 10% of the CO2
savings. The mitigation potential of CCS by sector is
illustrated in Fig. 5 for a total of 1,574.6 Mt CO2/year,
based on 2005 [17]. However, this potential is not
valued by the Government of Brazil for the actions of
development and domination of the CCS has been
headed primarily by Petrobras. Therefore, some
conditions are necessary:
 To establish a national regulatory framework for
this technology: the absence of an UNFCCC strategy
regarding the use of CCS to mitigate GHG emission
contributes for the no-decision of Brazilian
government;
 To approve the CDM methodology for CCS
technology and some national public fund to financing
Brazilian CCS projects in regulated and voluntary
Total = 1,574.6
Total = 1,574.6MtMt CO 2
CO2
CCS potential
Fig. 5 Brazilian inventory of anthropogenic emissions and removals of greenhouse gases in 2005. Index by Ref. [17].
31
carbon markets;
 To create public policy R&D mechanisms to
encourage more research and innovation in the CCS
technology.
If there is an incentive for the use of CCS in Brazil
(market for carbon credits or funds) exploratory
studies indicate a potential for CO2 mitigation, oil
provinces of the pre-salt and Recôncavo Basin,
according to the annual capacity of emissions, a total
of 26.4 Mt CO2/year is mitigated. Only Recôncavo
Basin, according Rocha (apud Costa in Ziliotto [7])
indicates a daily consumption of 14.5 kt CO2/day.
The Government of Brazil presented the GHG
reduction targets at COP 15 by means of “Nationally
Appropriate Mitigation Actions in Developing
Countries” (NAMAs). The targets were based on the
2020 projected total emissions of 2703 Mt CO2-eq
[20]. The contributions would be reduced by Sector:
 Land Use, 669 Mt;
 Agriculture and Livestok, 166 Mt;
 Power Generation, 207 Mt;
 Outhers, 10 Mt.
Overall reductions would be 1,052 Mt by 2020.
Exploratory studies indicate that CCS applied to the
Pre-salt and Recôncavo Basin has the potential to
mitigate CO2 per year around 25 Mt CO2/year based on
2020 the potential contribuition would be 275 Mt. This
would put Brazilian Governmental goals of GHG
reductions in an ambitious position, as shown in Fig. 6.
Change in land use and
forest
Burn fuel ‐ transport
Burn fuel ‐ industry
Burn fuel – others sectors
Industrial processes
Fugitive emissions
32 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
Fig. 6 Brazil’s national emissions and reduction potential under the Business-as-usual (scenario)-BAU and all reduction
scenarios between 1990 and 2020 (left) and the sensitivity analysis (right). Index by Ref. [21].
3.3 CCS’s Brazilian Players
The research and development of CCS started in
Brazil by Petrobras in the oil industry. According to
Lino [22], tests of CO2 injection in Buracica’s field in
the Reconcavo Basin/BA were started in May 1991.
After the oil industry other Brazilian productive sector
that started the research on CCS has been the coal
industry. The researches in this sector aim to study
using conventional and unconventional uses of coal
such as enhanced recovery of coal bed methane
(ECBM), methane contained in coal seam (CBM) and
hydrogen derived from coal seam via in situ coal
gasification (UCG).
Two important Brazilian institutions focused on the
CCS research and development are the Center for
Research and Development Leopoldo Américo
Miguez de Mello/Petrobras (CENPES) and the Center
for Excellence in Research on Carbon
Storage/PUC-RS (CEPAC). As for the capture of CO2,
is also worth mentioning the research activities
conducted by the University of Salvador (UNIFACS).
Other institutions that support research and
development of CCS in Brazil should be highlighted,
they are: Associação Brasileira do Carvão Mineral,
Associação Beneficente da Indústria Carbonífera de
Santa Catarina, COPELMI Minerações Ltda, Rede
Carvão, Instituto Ecoar and Instituto Ecoclima.
More broadly include:
 Climate Change: National Institute for Space
Research (INPE);
 Global Environmental Governance and CDM:
UFBA, World Policy Analysis Laboratory
(LABMUNDO).
Among private institutions, there is CEBDS
(Brazilian Business Council for Sustainable
Development), specifically the House Energy and
Thematic Climate Change. In the public the main
player is the Ministry of Science and Technology
(MCT) which is the competent national authority on
the UNFCCC about CDM.
Among the CCS Brazilian players the Petrobras is
leader in research and development actions. In the
inventory of Emissions of Greenhouse Gases of
Petrobras-base year 2008 presented to the Brazil GHG
Protocol Program of Petrobras the emissions increased
from 20.7 Mt CO2 in 1990 to 50.8 Mt CO2 in 2008. In
this report Petrobras announces that among the key
messages to combat the emissions, the company is
“investing aggressively in programs R & D to enable
new technologies, including renewables, energy
efficiency and CCGS (capture, sequestration, transport,
geological storage and monitoring of CO2), involving
several universities and research institutes” [23].
Petrobras is attentive to the emissions from the
processes of exploration, production, refining and
distribution of the oil industry, in addition, another
 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
important factor considered by the company is the
high presence of CO2 in the pre-salt hydrocarbon. The
characteristics of the pre-salt hydrocarbon are high
API, between 28-30°, high ratio of solution gas,
between 200 to 300 m3/m3 and presence of CO2 from
8% to 12% [24].
The following options to CCS in pre-salt are being
technically and economically evaluated: Enhanced Oil
Recovery (EOR) in the pre-salt areas; CO2 storage in
saline aquifers; EOR in heavy oil fields, in the Santos
Basin; CO2 storage in depleted gas fields; CO2 storage
in salt caverns, to be constructed in the cluster area;
CO2 transportation to shore and commercialization in
industrial plants (non-geological option). The
preferred option for the disposal of the CO2 rich
stream to be separated from the production gas seems
to be the reinjection in the hydrocarbon reservoir
where it came from, as there is a twofold benefit in
this strategy: providing some enhancement in the oil
recovery and ascertaining that the produced CO2 will
be effectively stored [24].
To anticipate environmental problems and also to
influence the global environmental governance (GEG)
climate, Petrobras has invested strongly in the domain
of CCS. One of the main results of these investments
is the CENPES and programs Proclima and PRO-CO2,
which are invested annually around $25 million.
Petrobras has also implemented the National Network
of Technology in Climate Change and CO2 Capture
and Sequestration. This network involves 13
institutions and from 2006 to 2009 it was invested
between $30 million in projects [24].
4. Conclusions
The CCS technologies have obtained international
notoriety for its high potential to mitigate GHG
emissions from stationary sources. These technologies
are being researched and developed a very high speed
in developed countries. The knowledge and domain of
these technologies is an important step for a country,
which can be considered a important player in the
33
forums about climate change. Evidence of this is in
the quantity of projects in the world, the resources
provided by developed countries and funding
mechanisms.
The Brazilian Government has not strongly
supported research and development of CCS
technologies, passing on the responsibility to private
sector. However, the private sector has almost
exclusively relied on investments of Petrobras and
coal industry. The gap of participation of the
Government of Brazil at a time when the international
scientific community, industries and governments of
developed countries are mobilizing to dominate the
technologies of CCS is critical.
Important aspects that need the effective
participation of the Government of Brazil to establish
a secure context for research and development of CCS
in Brazil should be defined as the regulatory
framework for CCS, forms of incentives, the
government’s definition of players, selection of
priority sectors, establishment of research networks,
among others.
This article provides an ample aspects to the large
scale use of CCS in Brazil and highlights the need for
future research on public perceptions in Brazil the use
of CCS, identification of regulatory gaps exist, the
state of the art of the CCS in Brazil and the CCS
interest of the Government and the private sector.
Specifically regarding the opportunities for new
researches about CCS technologies in Brazil, it is
still lacking some applied research, for example a pilot
project in commercial scale, in order to understand
and dominate all the technological stages of CCS from
the capture, transport, injection to the monitoring of
CO2 in geological reservoirs. Economic and financial
aspects of this kind of technology to mitigate GHG in
Brazil should also be subject of future researches.
References
[1] G. Porter, J.W. Brown, Global Environmental Politics,
2nd ed., Westview Press, Colorado, 1996.
[2] D. Helm, Climate-change policy: Why has so little been
34 Carbon Capture and Storage Technology: World Overview and Potential to Mitigate GHG in Brazil
achieved?, Oxford Review of Economic Policy 24
(2008) 211-238.
[3] S. Pacala, R. Socolow, Stabilization wedges: Solving the [15]
climate problem for the next 50 years with current
technologies, Science 305 (2004) 968.
[4] J. Meadowcroft, O. Langhelle, Caching the Carbon—The
Politics and Policy of Carbon Capture and Storage, [16]
Edward Elgar, Cheltenham, UK and Northampton, MA,
USA, 2009, p. 5.
[5] Energy technology perspectives 2006: Scenarios &
strategies to 2050 in support of the G8 plan of action,
International Energy Agency (IEA) Web site, [17]
http://www.iea.org/books.
[6] Intergovernamental Panel on Climate Change (IPCC),
IPCC special report on carbon dioxide capture and
storage, in: B. Metz, O. Davidson, H.C. de Coninck, M.
Loos, L.A. Meyer (Eds.), Cambridge University Press, [18]
Cambridge, 2005.
[7] M.A. Ziliotto, Mudanças climáticas, Sequestro e Mercado
de Carbono no Brasil, Curitiba, Brasil, 2009, pp. 335-339. [19]
(in Portuguese)
[8] Carbon dioxide enhanced oil recovery—Untapped
domestic energy supply and long term carbon storage
solution, National Energy Technology Laboratory (NETL)
Web site, http://netl.doe.gov. [20]
[9] F.Y. Inventário, National Injection Well 2000 [Online],
http://water.epa.gov/type/groundwater/uic/upload/2007_1
2_12_uic_UIC-Inventory.pdf.
[10] S. Brunsting, J. Desbarats, M. de Best-Waldhober, E. [21]
Duetschke, C. Oltra, P. Upham, et al., The Public and CCS:
The Importance of Communication and Participation in the
Context of Local Realities, Energy Procedia, 2010.
[11] J. Goldemberg, O caminho até Joanesburgo, Meio
ambiente no Século 21 (Energia), Rio de Janeiro, 2003, [22]
pp. 171-181. (in Portuguese)
[12] Views related to carbon dioxide capture and storage in
geological formation as a possible mitigation
technology—Submission from Parties, United Nations
Framwork Convention on Climate Change-UNFCCC,
2010, (Brazilian Submition on the possible inclusion of [23]
Carbon Dioxide Capture and Storage in Geological
Formation (CCS) in the Clean Development Mechanism),
Bonn. [24]
[13] ANSI GHG Valdation and Verification Body
Accreditation Scoping Policy, American National
Standarts Institute (ANSI), GHG-PR-706, 2009.
[14] WorleyParsons, Strategic Analysis of the Global Status of
Carbon Capture and Storage, Final report © 2009, Global
CCS Institute, Canberra, 2009.
International Energy Agency (IEA), Geologic storage of
carbon dioxide-staying safely underground, in:
Greenhouse Gas R&D Programme [Online],
http://www.ieagreen.org.uk.
Carbon Capture and Storage in the CDM [Online],
Organization for Economic Co-operation and
Development, 2007,
http://www.iea.org/textbase/papers/2007/CCS_in_CDM.p
df.
BRASIL. Lei nº 12.187, de 29 de dezembro de 2009.
Política Nacional de Mudança Climática, em 30 de
novembro de 2009. Diário Oficial [da República
Federativa do Brasil], Brasília, DF, 29 dez. Seção 1 -
Edição Extra, 2009, p. 109. (in portuguese)
Brasil, Ministério de Minas e Energia - MME e Empresa
de Pesquisa Energética-EPE. Plano Nacional de Energia
2030, Brasília: 2007, p. 324. (in portuguese)
C.X. Machado, J.M.M. Ketzer, A.J. Maraschin, G.C.
Rockett, G.F. Sbrissa, G. Caporale, et al., Construção de
um SIG para associação de fontes estacionárias e
reservatórios geológicos de CO2 no Brasil [Online], 2009,
http://www.ecoclima.org.br/.
Instituto de Pesquisa Econômica Aplicada (IPEA),
Perspectivas sobre as mudanças climáticas e seus
impactos na política brasileira, in: Comunicados do IPEA
No. 45, Brasília-DF, 2010, p. 15 (in portuguese).
S. Moltmann, M. Hagemann, K. Eisbrenner, N. Höhne,
W. Sterk, F. Mersmann, et al., Quantifying emission
reduction contributions by emerging economies, German
Federal Environment Agency Web site,
http://www.umweltdaten.de/publikationen.
U.R.A. Lino, Case history of breaking a paradigm:
Improvement of an immiscible gas-injection project in
buracica field by water injection at the gas/oil contact,
Presented at: Spe Latin American and Caribbean
Petroleum Engineering Conference [Online],
http://www.spe.org.
PETROBRAS, Programa Brasileiro GHG Protocol 2009,
Inventário das Emissões de Gases de Efeito Estufa-2008,
2009. (in portuguese)
A.S. Almeida, P.S. Rocha, S. de T. Lima, A.C.C. Pinto,
C.C. Branco, M.C. Salomão, A study on the potential dor
CCGS in the pre-salt cluster of Santos Basin: The Tupi
pilot application, in: Rio Oil & Gas Expo and Conference
2010, 2010.
Journal of Environmental Science and Engineering A 1 (2012) 35-42
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Continuous Reaction Crystallization of Struvite from

Water Solutions of Phosphates(V) in Presence of Iron(II)
Ions

N. Hutnik1, K. Piotrowski2, B. Wierzbowska1 and A. Matynia1

1. Faculty of Chemistry, Wroclaw University of Technology, Wroclaw 50-370, Poland
2. Department of Chemical & Process Engineering, Silesian University of Technology, Gliwice 44-101, Poland

Received: May 9, 2011 / Accepted: July 29, 2011 / Published: January 20, 2012.

Abstract: The experimental results concerning removal of phosphate(V) ions from their diluted water solutions (1.0 mass % of PO43-)
with the use of magnesium and ammonium ions in stoichiometric conditions, in the presence of iron(II) ions, are presented and
discussed. Laboratory tests were carried out in a continuous DT (Draft Tube) MSMPR type crystallizer with propeller mixer. The
effects of iron(II) ions concentration (from 1 × 10-4 to 1 × 10-3 mass %) and pH of reaction crystallization environment (pH = 9-11) on
the size distribution of struvite crystals, their habits and chemical composition of mother solution/solid phase were determined
experimentally and discussed. It was concluded that increase in iron(II) ions concentration in a feed solution influences mean crystal
size of the product disadvantageously. Increase in pH value in a process system produces similar adverse effects. Practical complete
precipitation of iron(II) ions was confirmed analytically-their concentration in the outlet mother solution was 0.05-0.07 mg/kg. The
process resulted in formation of a binary solid mixture composed of struvite crystals and iron hydroxide particles. Process efficiency of
phosphate(V) ions removal in the presence of iron(II) ions in the proposed technological/apparatus conditions reached the level of ca.
98%.

Key words: Phosphorus recycling, phosphate(V) ions, struvite, reaction crystallization, continuous DT MSMPR crystallizer, iron(II)
ions, iron hydroxide, agglomeration.

Nomenclature manure, etc. is their chemical precipitation as sparingly

CSD Crystal size distribution soluble calcium (hydroxyapatite, HAP, brushite,
CV Coefficient of variation (of crystal size) DCPD) [1] or magnesium (struvite, MAP) salts [2].
DCPD Dicalcium phosphate dihydrate, brushite, This last compound, MgNH4PO4·6H2O, can be even
CaHPO4·2H2O
utilized directly in agricultural industry as a valuable
DT Draft tube (crystallizer configuration)
HAP Hydroxyapatite, Ca5(PO4)3OH mineral fertilizer [3, 4]. All phenomena accompanying
MAP Magnesium ammonium phosphate hexahydrate, formation of useful forms of recovered phosphorus
struvite, MgNH4PO4·6H2O from the diluted phosphate(V) solutions are generally
MSMPR Mixed suspension mixed product removal
named as phosphorus recycling processes [2, 5]. The
(crystallizer configuration)
examples of more or less successful tests of
1. Introduction technological and apparatus solutions concerning the
process of struvite recovery in laboratory, pilot and
One of the possibilities of phosphates(V) recovery
technical scale are described in Refs. [4-9]. These can
from municipal and/or industrial liquid wastes, liquid
be also found in the review works of Doyle and Parsons
Corresponding author: K. Piotrowski, Ph.D., main research [5, 10], as well as of Le Corre et al. [11].
fields: reaction crystallization kinetics, crystallizer design, Final effects of the controlled precipitation and
process modeling. E-mail: krzysztof.piotrowski@polsl.pl.
36 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence
of Iron(Ii) Ions
crystallization of struvite-crystal population of
determined chemical purity and crystal size
distribution (CSD), produced with satisfactory yield,
depend on many parameters determining the process
environment, on these parameter values, as well as on
the constructional solutions applied [2, 5, 10-14]. One
of the factors influencing the process course, thus the
quality of the crystal product, are impurities present in
waste solutions, mainly metal ions, sulfates(VI),
nitrates(V), fluorides, fluosilicates and others [2, 5].
They inhibit or catalyze the course of struvite
precipitation, influence nucleation and crystal growth
rates, modify crystal shape and kinetics of eventual
agglomeration processes within the individual crystals
[15-17]. Some of the metal ions in alkaline
environment of struvite precipitation can also
co-precipitate in a form of sparingly soluble
hydroxides or phosphates, what essentially deteriorates
chemical composition of the product thus the
possibilities of its further processing and final
application [18].
One of the impurities present in liquid wastes,
among others in these derived from industrial
production of phosphoric acid [19], are iron(II) ions.
The experimental results concerning influence of
presence and concentration of iron(II) ions on the final
results of reaction crystallization process of struvite in
a continuous DT MSMPR (Draft Tube, Mixed
Suspension Mixed Product Removal) type crystallizer
with propeller mixer are presented below. The process
yield based on difference between final concentration
of PO43- ions in a postprocessed mother solution and its
initial concentration in a feed was determined.
Statistical parameters of product crystal size
distribution (Lm, L50, CV), crystal shape and
agglomeration effects were estimated. Analytical
identification of separated solid phases was done.
Process was carried out in a constant temperature 298
K. Mean residence time of suspension in a crystallizer
working volume, , was assumed to be 900 s.
Laboratory crystallizer was fed with water solution
containing phosphate(V) ions of concentration 1.0
mass %, magnesium and ammonium ions in
stoichiometric (in respect to PO43-) proportions, as well
as with iron(II) ions (in a chemical form of iron(II)
chloride) which concentration varied from 1 × 10-4 to 1
× 10-3 mass % (1-10 mg/kg). Main set of experiments
was done at pH = 9, while for two selected
concentrations of Fe2+ ions in a feed solution, 2 × 10-4
and 5 × 10-4 mass %, additional tests at higher pH (10
and 11) were also carried out.
2. Experimental Setup and Procedure
Scheme of experimental research stand for the tests
concerning reaction crystallization of struvite in the
presence of iron(II) ions is shown in Fig. 1.
Automatic operation and acquisition of
measurement data were carried out with the use of PC
computer. For the precise control of measurement
stand working in steady-state conditions the
professional IKA labworldsoft system was used.
Working volume of a laboratory DT MSMPR
crystallizer was Vw = 0.6 dm3. It was a cylindrical glass
tank (D = 120 mm, H = 123 mm), in bottom part
equipped with external heating/cooling coil made with
stainless steel, providing constant process temperature.
In the central axis of apparatus a cylindrical circulation
profile (draft tube: ddt = 57 mm, hdt = 53 mm) was
located, inside which a three paddle propeller mixer
(dm = 55 mm) was running. Localizations of inlet
positions of ionic reagents mixture, pH correction
agent as well as the place of withdrawal of product
crystal suspension are indicated in Fig. 1. Crystallizer
was continuously fed with water solution of
magnesium chloride, MgCl2, ammonium
dihydrogenphosphate, NH4H2PO4 and iron(II) chloride,
FeCl2. Feed mixture was continuously introduced
inside the circulation profile interior (mixer revolution
number: 6.6 ± 0.1 1/s; downward circulation of
suspension). Between the crystallizer body and draft
tube element (upward circulation of suspension) water
solution of sodium hydroxide, NaOH (of concentration
 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence 37
of Iron(Ii) Ions

5
NH4H2PO4 +
MgCl2 + 7 8
FeCl2 NaOH
3 solution solution pH
M
6 T
4 9
10 11

product
12
1
2

Fig. 1 Scheme of a laboratory experimental stand with a continuous DT MSMPR crystallizer with propeller mixer for
reaction crystallization of struvite, 1-DT MSMPR crystallizer with internal circulation of suspension, 2-thermostat, 3-PC
computer, 4-reservoir of a feed mixture: water solution of NH4H2PO4, MgCl2 and FeCl2, 5-dosing pump of a feed mixture,
6-alkaline agent tank: water solution of NaOH, 7-dosing pump of NaOH solution, 8-pump for isokinetic removal of crystal
suspension from the crystallizer working volume, 9-container for product crystal suspension, 10, 11, 12-electronic balances,
M-mixer speed control/adjustment, pH-control/adjustment of pH, T-temperature control/adjustment.

20 mass %) was dosed in required proportions
providing the assumed and strictly controlled pH of the
phosphates(V) precipitation environment.
Main experimental tests were carried out in process
temperature 298 K assuming pH 9 and mean residence
time of suspension in a crystallizer working volume
900 s. Concentrations of individual reagents in a feed
mixture were assumed constant: [PO43-]RM = 1.0
mass %, [Mg2+]RM = 0.256 mass % and [NH4+]RM =
0.190 mass % (resulting in molar ratio between the
main reagents as 1:1:1). Concentration of iron(II) ions
was changed from 1 × 10-4 to 1 × 10-3 mass %. For the
two selected concentrations of Fe2+ ions, 2 × 10-4 and 5
× 10-4 mass %, additional tests at pH raised up to 10
and 11 were also done (see Table 1). After stabilization
of the assumed process parameter values in crystallizer
the process in a steady state was carried out through 5.
After this time the process was terminated and crystal
product characteristics were determined using standard
analytical methods. The characteristics covered:
concentration of solid phase in a product crystal
suspension (MT), size distribution (CSD) of crystal
phase (solid particle analyzer COULTER LS-230),
chemical composition of mother solution and solid
phase, as well as crystal shape (analysis of images from
scanning electron microscope JEOL JSM 5800LV).
3. Results and Discussion
The test results are presented in Table 1. It can be
concluded that presence of iron(II) ions in a feed
solution within the 1 × 10-4-1 × 10-3 mass %
concentration range (pH 9) influences product CSD
disadvantageously. Both statistical parameters of size
distribution (Lm and L50) decrease. Mean crystal size Lm
decreases from 36.5 to 28.6 m (by more than 20%),
median size L50 (for the undersize/oversize crystal
fraction equals 50 mass %) from 30.9 to 23.9 m (by
more than 22%). Moreover, the 10 time increase in
concentration of iron(II) ions in a feed results also in a
decrease of the size homogeneity within the struvite
product crystals. The CV coefficient values, regarded as
a useful form of quantification of inhomogeneity within
38 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence
of Iron(Ii) Ions

Table 1 Chemical analyses results and statistical parameter values of crystal size distributions of struvite produced in a
continuous reaction crystallization process of phosphate(V) ions precipitation by means of magnesium and ammonium ions
addition (in the stoichiometric proportions) in the presence of co-precipitating iron(II) ions.
Concentration of Fe2+ ions Concentration of Fe(total) Crystal characteristics
pH
in a feed solution (mass %) in a product (mg/kg) Lm (μm) L50 (μm) CV (%)
9 1 × 10-4 40 36.5 30.9 73.0
9 2 × 10-4 82 34.9 30.5 72.8
9 3 × 10-4 125 34.0 25.8 89.8
9 5 × 10-4 205 33.0 27.4 81.8
9 1 × 10-3 425 28.6 23.9 81.0
10 2 × 10-4 84 29.1 25.6 75.1
11 2 × 10-4 85 18.2 13.2 88.9
10 5 × 10-4 208 23.8 17.8 85.5
11 5 × 10-4 215 16.1 11.6 88.2
Concentration of main reagents in a feed solution: 1.0 mass % of PO43-, 0.256 mass % of Mg2+, 0.190 mass % of NH4+.
Mean residence time of crystal suspension in working volume of a crystallizer:  = 900 s.
Mean concentration of product crystals in suspension: MT = 23.9 ± 0.4 kg of struvite/m3.
Concentration of iron(II) ions in mother solution (as total Fe): 0.05-0.07 mg/kg.
Concentration of phosphate(V) ions in mother solution: 0.013  0.002 mass %.
Process temperature: 298 K.

the crystal sizes, increased from 73% to above 80%.
With the increase in pH of struvite reaction
crystallization process environment product crystal
size also decreased. Increase in pH from 9 to 11,
assuming concentration of Fe2+ ions in a feed solution
to be 2 × 10-4 mass %, resulted in nearly two time
decrease in mean crystal size, Lm (34.9→18.2 m) and
more than two time decline (33.0→16.1 m) while
assuming 2.5-time higher their concentration in the
feed (5 × 10-4 mass %). This trend is also clearly
observable analyzing the values of median crystal size,
L50 (see Table 1). Increase in pH does not also favor the
homogeneity within the crystal product population. At
pH 11 coefficient of crystal size variation CV reached
the 88%-89%. With the increase in pH solubility of
struvite decreases, thus its precipitation potential
increases [2, 5]. Also induction time, indispensable for
the nucleation process [20], considerably shortens. All
these result in increase in population density of nuclei,
externally observed as the clear shift of mean (or mode)
crystal size towards lower values [12].
In Fig. 2, there are presented some exemplary
volumetric (mass) CSDs of solid product,
corresponding to pH 9, 10 and 11 ( = 900 s,
concentration of iron(II) ions in a feed solution 5 × 10-4
mass %). For very small crystals (0.4-4 m) produced
at pH 9 (Fig. 2a) some clear disturbance in a smooth
course of differential distribution curve (local
maximum) is observed. Within this size range one
deals with unexpected excess number of very small
particles, what is connected with the presence in a
product crystals population additional type(s) of
particles other than struvite crystals. This was also
confirmed by physicochemical analytical tests (see
Table 1). These identified species are mainly particles
of iron(III) hydroxide, Fe(OH)3 (which solubility
product in temperature of 298 K is Ksp = 1.1 × 10-36,
pKsp = 35.96), as well as, however in a considerable
lower proportion, iron(II) hydroxide, Fe(OH)2 (Ksp =
1.8 × 10-14, pKsp = 13.74). In this way in the presented
process conditions heterogeneous population in respect
to chemical composition of solid phase was obtained as
a final product. With the increase in pH local maximum
observed in a differential distribution curve becomes
gradually less and less clear (pH 10, Fig. 2b), and
finally even disappears (pH 11, Fig. 2c). It is connected
with a general increase in the number of the smallest
struvite crystals within the product population which
 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence
of Iron(Ii) Ions
(a)
(b)
(c)
Fig. 2 Influence of reaction crystallization process pH on
differential (left scale) and cumulative (right scale)
volumetric (mass) size distributions of struvite crystals
produced in a continuous DT MSMPR crystallizer in the
presence of co-precipitating iron(II) ions (5 × 10-4 mass % of
Fe2+ ions in a feed solution): (a) pH 9; (b) pH 10; (c) pH 11
(see Table 1 for details).
contribute to this size fraction. At pH 11 (and possibly
higher) the struvite crystals of sizes below 4 m are
practically indistinguishable from other particle types
(based on CSD analysis only). Crystal mode size (Ld,
corresponding to maximum in differential volumetric
(mass) size distribution) gradually shifts towards lower
39
sizes. Thus in a product population both the number
and characteristic size of the largest crystals decrease.
The largest size of struvite crystals produced at pH 9 is
ca. 160 m (Fig. 2a), at pH 10 it is ca. 130 m (Fig. 2b),
while at pH 11 it is only ca. 85 m (Fig. 2c).
Simultaneously one can observe systematic increase in
the fraction of particles of the smallest sizes. At pH 9
fraction of crystals of sizes below 10 m was ca. 22%,
while at pH 11 this fraction grew up to ca. 44%, thus
two-time increase was detected (compare Figs. 2a and
2c). In result mean size of product crystal population
decreased considerably from 33.0 to 16.1 m (see
Table 1 for details).
In Fig. 3 there are presented the images of product
crystals, which corresponding CSDs are presented in
Fig. 2. Smaller sizes of struvite crystals produced at
higher pH values of reaction crystallization process
environment are clearly visible. On the other hand it
can be noticed that struvite crystals habit was
practically unmodified. On the basis of planimetric
measurement results covering 50 individual crystals
selected randomly from each crystal population it was
calculated that their length La to width Lb ratio is of
average value 5.0  0.2. This proportion is stable
irrespective of pH and concentration of iron(II) ions in
a feed solution. On the surface of MgNH4PO4·6H2O
crystals one is able to observe the effects of surface
aggregation/agglomeration of distinctly smaller
particles of iron hydroxide (promoted especially by the
occurrence of irregularities within the crystal surface,
called surface molecular roughness).
Besides, the iron hydroxide particles are also
observable in a whole crystal population, the most
often in agglomerate forms. Contrary, for the struvite
crystals it is able to observe less than moderate effects
of agglomeration, especially within the crystals of
larger sizes. It can be thus concluded that the
advantageous process conditions were created in a DT
MSMPR crystallizer, effectively limiting excessive
sticking and aggregation/agglomeration within the
crystals of a main product component.
40 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence
of Iron(Ii) Ions

100 m From the scanning electron microscope images of

(a)
100 m
(b)
100 m
(c)
Fig. 3 Scanning electron microscope images of struvite
crystals produced in a continuous DT MSMPR crystallizer
in the presence of co-recipitating iron(II) ions (5 × 10-4
mass % of Fe2+ ions in a feed solution): (a) pH 9; (b) pH 10;
(c) pH 11. Magnification 500. Mean residence time of
crystal suspension in a crystallizer working volume: 900 s.
crystal product it also results that within the struvite
crystals some characteristic in shape, tubular forms are
present. According to Ref. [21], the struvite crystals
show this type of habit especially at pH 7. In the
presence of iron(II) ions tubular crystals prevailed in
struvite product. However, majority of them
demonstrated, especially crystals of larger sizes, some
distinct cracks and fractures along longitudinal axis,
many surface defects and deformed ends, indicating
occurrence of considerable stresses within their
internal structure. Crystals formed in result of pure
molecular growth processes (e.g. dendritic growth) are
theoretically especially susceptible on the action of
shearing forces induced by turbulent flow of mother
solution, as well as potentially damageable in result of
contact secondary nucleation processes (facilitating by
intensive mixing, relatively high concentration of
crystals in the circulated suspension, etc.).
In Table 1 there are presented the analytically
determined concentrations of phosphate(V) and iron(II)
ions in mother solution, as well as total iron content in a
final solid product (Fe(total)). In practice it can be
assumed that concentration of phosphate(V) ions in
mother solution keeps the relative constant and low
level of 0.013  0.002 mass % while concentration of
iron(II) ions in some environment (presented as total Fe)
decreased to a level below 0.07 mg/kg. One thus
observes their nearly total precipitation in a chemical
form of iron hydroxides, which appeared in the solid
phase in result of co-recipitation processes in
concentration range varied from 40 to 425 mg/kg (on
the total iron basis).
Reaction crystallization process of struvite was
carried out assuming relatively short mean residence
times of product crystal suspensions in a crystallizer
working volume, thus at potentially high levels of
working supersaturation in solution. However, one is
not able to credibly determine (even as an
approximation) this supersaturation level since the
value of solubility product of struvite in the discussed
 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence
of Iron(Ii) Ions
reaction crystallization process system is not exactly
known. This value depends on many process
parameters (e.g. mixing intensity responsible for the
induction and distribution of turbulence), presence and
concentration of third substances (iron(II) ions, as well
as chloride and sodium ions) but, first of all, on pH of
the process environment. More detailed insight into
these complex ionic equilibria, as well as into influence
of pH on the struvite solubility product value were
presented elsewhere [12, 18, 22, 23].
4. Conclusions
Precipitation of phosphate(V) ions with the use of
magnesium and ammonium ions, in presence of iron(II)
ions, results in separation from feed solution not only
struvite crystals, but also co-precipitated particles of
various iron hydroxides. Contaminated product is thus
formed, what decreases its market value and effectively
limits its eventual application in agricultural industry.
The process was carried out in temperature 298 K, in
stoichiometric conditions, at pH 9-11. Mean residence
time of product crystal suspension in a crystallizer
working volume was assumed to be 900 s.
Concentration of phosphate(V) ions in a feed solution
was constant (1.0 mass %), as well as
stoichiometrically correlated concentrations of
magnesium and ammonium ions (see Table 1), whereas
concentration of iron(II) ions was changed from 1 ×
10-4 to 1 × 10-3mass % (1-10 mg/kg). It was observed
that with the increase in concentration of iron(II) ions
in a system mean crystal size decreased by more than
20% (from 36.5 to 28.6 m at pH 9). Homogeneity
within the population of product crystals decreased, as
well. Increase in pH of reaction crystallization process
environment additionally influences crystal product
quality disadvantageously. Crystals of smaller sizes
(Lm = 33.0 → 16.1 m at pH 9 → 11, for the iron(II)
ions concentration in a feed solution 5 × 10-4 mass %)
and of lower homogeneity (CV = 81.8 → 88.2%) were
produced. Besides, presence of iron(II) ions favored
crystallization of struvite in a characteristic form of
41
tubular crystals. These crystals have cracks and
fractures along longitudinal axis and on their surfaces,
as well as clearly observed shredded ends. Total
removal of mother solution from the interiors of such
tubular forms during their filtering/washing operations
is practically impossible. It was also observed that
particles of iron hydroxides, simultaneously
co-recipitated with the struvite product crystals,
characterized by very small sizes, even below 1 m.
Moreover, they formed surface agglomerates on the
larger, parent struvite crystals, as well as independent
agglomerate structures. Presence of the last forms in
the product crystal suspension makes phase separation
very difficult and elongates process time necessary for
effective filtration and washing of the resulting
sediments. Iron hydroxides content in a final solid
product, recalculated as total Fe, varied from 40 to 425
mg/kg depending on concentration of iron(II) ions in a
feed. In practice total precipitation of iron(II) ions was
observed since their concentration in a postprocessed
mother solution was below 0.07 mg/kg. Concentration
of phosphate(V) ions in this solution was 0.013  0.002
mass %, thus it decreased by ca. 98% in respect to their
initial concentration in a feed inlet point. This effect
can be regarded as a good result of their removal in a
continuous reaction crystallization process using the
described constructional/technological alternative.
Acknowledgments
The work was financed by Ministry of Science and
Higher Education of Poland in the period of 2006-2009
and 2008-2011 under grants No. R05 053 01 and No.
N209 010834.
Physicochemical analyses of crystal phase and
mother solution were done in Institute of Inorganic
Chemistry in Gliwice, Poland.
References
[1] E. Valsami-Jones, Recent scientific and technical
developments: Calcium phosphate precipitation, CEEP
Scope Newslett. 41 (2001) 8-15.
[2] S.A. Parsons, Recent scientific and technical
42 Continuous Reaction Crystallization of Struvite from Water Solutions of Phosphates(V) in Presence
of Iron(Ii) Ions
developments: Sruvite precipitation, CEEP Scope
Newslett. 41 (2001) 15-22.
[3] G. Bridger, Fertiliser value of struvite, CEEP Scope
Newslett. 43 (2001) 3-4.
[4] L.E. de-Bashan, Y. Bashan, Recent advances in removing
phosphorus from wastewater and its future use as fertilizer,
Wat. Res. 38 (2004) 4222-4246.
[5] J.D. Doyle, S.A. Parsons, Struvite formation, control and
recovery, Wat. Res. 36 (2002) 3925-3940.
[6] R.D. Schuiling, A. Andrade, Recovery of struvite from
calf manure, Environ. Technol. 20 (1999) 765-768.
[7] J. Greaves, P. Hobbs, D. Chadwick, P. Haygarth,
Prospects for the recovery of phosphorus from animal
manures: A review, Environ. Technol. 20 (1999) 697-708.
[8] D. Donnert, M. Salecker, Elimination of phosphorus from
wastewater by crystallization, Environ. Technol. 20 (1999)
735-742.
[9] A. Giesen, Crystallisation process enables environmental
friendly phosphate removal at low costs, Environ. Technol.
20 (1999) 769-775.
[10] S.A. Parsons, J.D. Doyle, Struvite scale formation and
control, Wat. Sci. Technol. 49 (2004) 177-182.
[11] K.S. Le Corre, E. Valsami-Jones, P. Hobbs, S.A. Parsons,
Phosphorus recovery from wastewater by struvite
crystallization: A review, Crit. Rev. Environ. Sci. Technol.
39 (2009) 433-477.
[12] A. Matynia, J. Koralewska, K. Piotrowski, B.
Wierzbowska, The influence of process parameters on
struvite continuous crystallization kinetics, Chem. Eng.
Comm. 193 (2006) 160-176.
[13] J. Koralewska, K. Piotrowski, B. Wierzbowska, A.
Matynia, Kinetics of reaction-crystallization of struvite in
the continuous draft tube magma type
crystallizers-influence of different internal hydrodynamics,
Chinese J. Chem. Eng. 17 (2009) 330-339.
[14] J. Koralewska, N. Hutnik, K. Piotrowski, B. Wierzbowska,
A. Matynia, Effect of reactive crystallization parameters
on the quality of struvite crystals produced in a continuous
draft tube magma crystallizer with a jet pump fed with
recirculated mother solution, Przem. Chem. 88 (2009)
472-476. (in Polish)
[15] N. Hutnik, K. Piotrowski, B. Wierzbowska, A. Matynia,
Influence of sulphate(VI) and nitrate(V) ions on the
quality of struvite crystals in a continuous reaction
crystallization process for removing phosphates from
diluted water solutions, Progr. Environ. Sci. Technol. 2
(2009) 1044-1049.
[16] N. Hutnik, B. Wierzbowska, A. Matynia, K. Piotrowski,
Influence of potassium ions on the quality of struvite
crystals produced in continuous reaction crystallization
process, Inz. Ap. Chem. 48 (4) (2009) 54-55. (in
Polish)
[17] N. Hutnik, B. Wierzbowska, A. Matynia, K. Piotrowski,
Influence of zinc ions on the quality of struvite crystals
produced in continuous crystallization process from
diluted water solutions, Inz. Ap. Chem. 48 (5) (2009)
38-39. (in Polish)
[18] N. Hutnik, B. Wierzbowska, A. Matynia, K. Piotrowski, J.
Gluzinska, Influence of aluminium ions on the quality of
struvite crystals produced in a continuous mode from
the diluted solutions, Chemik 61 (2008) 505-508. (in
Polish)
[19] P. Becker, Phosphates and phosphoric acid raw materials,
in: Technology and Economics of the Wet Process, Marcel
Dekker, New York, 1999.
[20] F. Abbona, R. Boistelle, H.E. Lundager, Crystallization of
two magnesium phosphates, struvite and newberyite:
Effect of pH and concentration, J. Cryst. Growth 57 (1982)
6-14.
[21] M.M. Seckler, M.L.J. van Leeuwen, O.S.L. Bruinsma,
G.M. van Rosmalen, Phosphate removal in a fluidized
bed-II. Process optimization, Wat. Res. 30 (1996)
1589-1596.
[22] K.N. Ohlinger, T.M. Young, E.D. Schroeder, Predicting
struvite formation in digestion, Wat. Res. 32 (1998)
3607-3614.
[23] K.N. Ohlinger, Struvite precipitation kinetics, Scope
Newslett. 36 (2000) 6-8.
Journal of Environmental Science and Engineering A 1 (2012) 43-61
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Calculation of Surface Tracer Concentration by Set of

Hydrodynamic Weather Forecast Model and Lagrangian
Model of Atmospheric Dispersion

Konstantin G. Rubinstein1, Elena M. Nabokova1, Roman Y. Ignatov1, Maria M. Smirnova1, Rafael V. Arutunyan2,
Vladimir N. Semyonov2, Olga S. Sorokovikova2, Aleksey V. Fokin2 and Dmitry A. Pripachkin2
1. Hydrometeorological Centre of Russian Federation, Moscow 123242, Russia
2. Nuclear Safety Institute, Russian Academy of Sciences, Moscow 115191, Russia

Received: May 16, 2011 / Accepted: June 23, 2011 / Published: January 20, 2012.

Abstract: This study attempts to test, configure and improve a combined set of models comprising a regional hydrodynamic model
WRF ARW (USA, Weather Research and Forecast) and a model of radionuclide transport of the NOSTRODAMUS code. Performance
of the hydrodynamic model has been evaluated by comparing its results with the measurements from a meteorological tower; the set of
models has been evaluated using actual measurements of atmospheric tracer concentrations in field experiments performed in
1983-1985 near Karlsruhe, Germany. The paper includes a short description of the versions of the hydrodynamic forecast model and
the radionuclide transport model, which have been prepared for this study. The results of statistical analysis of various numerical
experiments involving the set of models are also present in the paper. Some experiments using data from various reanalyses as initial
and boundary conditions of the model have been performed and analyzed. It has been demonstrated that errors in vertical distribution of
model wind are extremely sensitive to the type of parameterization of the model surface and boundary layers. The statistical
characteristics of comparison of measured and calculated values of gas concentration are given. Testing of the set of models has
demonstrated that the combination of the regional forecast model with high spatial resolution together with the transport model of the
NOSTRODAMUS code allows to obtain distribution of surface gas concentrations with sufficient precision.

Key words: Radionuclide transport, regional hydrodynamic modeling, turbulent dispersion, meteorological forecast.

1. Introduction combined with radionuclide transport model and

description of horizontal and vertical turbulence in the
At present a significant increase in the number of
boundary layer and in free atmosphere.
nuclear power plants is expected in Russia. The need to
The objective of such systems of combined models
assure safe operation of the plants leads to the necessity
is to provide real-time data support for justified
of creating and developing mathematical models of
decision-making regarding actions in case of nuclear
operative prediction of emergency radioactive
plant accidents (timely warning of public and nuclear
discharge dissemination for tenths and hundreds of
power plant employees, implementation of protective
kilometers in the atmosphere in case of accidents at
actions and evacuation of people) [1-10].
nuclear plants. Such models are to be provided with
This study attempts to test and improve a combined
prognostic fields of meteorological parameters
set of models comprising a regional hydrodynamic
calculated using specific versions of regional
model WRF ARW (USA, Weather Research and
hydrodynamic models with high spatial resolution,
Forecast) [http://www. mmm.
Corresponding author: Konstantin G. Rubinstein, Ph.D., ucar.edu/wrf/users/download/get_source2.html] and a
main research fields: climate and mesoscale modeling. E-mail: model of radionuclide transport taken from the
k.g.rubin@googlemail.com.
44 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
NOSTRODAMUS code [2]. Performance of the
hydrodynamic model has been evaluated by comparing
its results with the measurements from a
meteorological tower; and the set of models has been
evaluated based on actual measurements of
atmospheric tracer concentrations obtained in field
experiments. The field experiments were performed in
1983-1985 near Karlsruhe, Germany [11] (see
description of the experiments in section 4). There
were four experiments performed in total, in which
surface gas concentration was measured at different
distances from the source. Details of the conditions and
results of the field experiments are described in the
text.
The paper includes a short description of the version
of the hydrodynamic forecast model prepared for this
work and the radionuclide transport model of the
NOSTRODAMUS code [2].
Results of statistical analysis of different numerical
experiments performed with the set of models are
presented.
2. Version of the Hydrodynamic Forecast
Model
A configuration of the regional non-hydrostatic
WRF-ARW model with a spatial resolution of 2 km
was created for hydrodynamic prediction of
meteorological fields in the South of Germany, near
Karlsruhe.
The area of forecast had the dimensions of 165
points by 145 points, which comprised 330 km by 190
km. The atmosphere was divided vertically into 24
layers up to 50 gPa layer. The lower 15 layers were
chosen within the boundary layer.
The need for such a high spatial resolution here was
due to that the valley near Karlsruhe also has a
complicated topography and it was assumed that an at
most precise description of the underlying surface
could improve the calculated results.
The WRF ARW model includes a large library of
procedures, describing physical processes in free
atmosphere, its boundary layer, the bottom surface and
in the soil. Availability of different procedures
describing the same processes allows choosing the
most adequate combination, which would be optimal for
a given terrain and would provide the best description of
the processes in the atmosphere for specific purpose,
local conditions and spatial scale, which are determined
by horizontal and vertical resolution.
The time step in the experiments was 10 seconds.
One may get an impression of the area of forecast by
looking at Fig. 1a, which demonstrates the terrain relief
used in the model.
In field experiments the tracer source was located on
a tower with the following coordinates: in the latitude
49.05° North and in the longitude 8.39° East. This
point was selected as the center of the prognostic field.
Field experiments took a short interval of
time—about 3 or 4 hours. Determination of wind fields,
which would be close to real ones, on a complicated
terrain in a short time frame was an extremely
important and difficult task. These fields were to a
great extent determined by initial data. To evaluate
their impact on simulation results experiments with
initial data and boundary conditions were carried out
for a model of three datasets:
(1) NCAR/DOE reanalysis [12], USA, with a spatial
resolution of 2.5 degrees of latitude and longitude on
18 standard isobaric surfaces from 1,000 to 10 gPa;
(2) ECMWF reanalysis [13] with a spatial resolution
of 2.5 degrees of latitude and longitude on 23 standard
isobaric surfaces from 1000 to 1 gPa;
(3) JRA reanalysis [14] by the Japanese
meteorological office with a spatial resolution of 1.25
degrees of latitude and longitude on 25 levels from
1,000 to 10 gPa.
Vertical profiles of the model wind obtained with
different initial data were compared to the values
measured on a tower.
This paper also contains results obtained in the
experiments with different parameterizations of the
surface and planetary boundary atmosphere layers in
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model 45
and Lagrangian Model of Atmospheric Dispersion

(а) (b)
Fig. 1 (а) Relief of the Karlsruhe test site. (b) Location of the mast with the source and the tower for meteorological
measurements (the cross with coordinates corresponds to a point with the coordinates of 49.05° N and 8.39° E) and the network
of stations, where the surface gas concentration was measured during filed experiments.

the hydrodynamic model. A short description of these
techniques follows. Parameterizations of other physical
process have not been changed in the experiments.
A. Parameterizations of processes in the surface
layer
(1) Surface layer parameterization based on the
theory of similarity. This procedure uses stability
functions described in Ref. [15] for calculation of
exchange factors for heat, moisture and momentum
fluxes. Convective rate calculated according to Ref.
[15] is used to obtain a realistic description of surface
heat and moisture fluxes. The ratio between the
roughness parameter and the friction velocity above the
water surface is derived from the Charnock formula.
According to Ref. [16], 4 modes of stability may be
distinguished.
(2) Another model of the surface layer also based on
the theory of similarity. It is based on a scheme of the
surface layer described in Refs. [17, 18]. It includes
parameterization of a viscous sublayer. The viscous
sublayer above the water surface is explicitly
parameterized as described in Ref. [18]. The impact of
the viscous sublayer overland is accounted for by using
variable roughness according to the method of
Zilitinkevich [19] when calculating heat and moisture
fluxes. Fluxes on the surface are calculated using
iteration technique.
B. Parameterization of soil processes
(1) Land surface model is based on work of Chen et
al. [20]. This parameterization allows predicting soil
moisture and temperature in four layers (10, 30, 60 and
100 cm thick) and a water equivalent of snow. It also
allows obtaining surface and subsurface run-off. The
parameterization requires information regarding soil
type and vegetation to determine evapotranspiration.
Characteristics of such underlying surfaces as ice and
sea are also taken into account. The scheme includes
prediction of snow thickness and soil freezing. Soil
water is also a decision variable. Soil humidity is
defined as a sum of soil moisture and ice in the soil.
(2) Land surface model is based on paper of Xiu et al.
[21]. The main ideas in this model are adopted from a
rather popular parameterization, called ISBA
(Interactions between Soil, Biosphere, and
46 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion

Atmosphere). Changes have been introduced to the on solving Eq. (1):

concept of the evapotranspiration process. The soil
   C z 
3
C 
consists of two layers: a surface layer of 1 cm thickness  K
 c    c   ( w  c  ) h    (1)
t  z    z   h  
and a layer of roots down to 1 m. 5 equations are solved:
2 equations for temperature in 2 layers of the soil, 2 where С is a prognostic variable composed of (u, v,
equations for humidity in 2 layers of the soil and an  , q ), К is a corresponding turbulence coefficient, and
с
equation for transfer of moisture by vegetation. The c is the correction to the local gradient, which accounts
relaxation time is 24 hours for the surface layer and 10 for the contribution of large horizontal whirlwinds to the
days for the meter-deep layer. Evaporation is defined total flux, z is the height above the surface, h is the
as a sum of evaporation from the surface layer, the height of the boundary layer, ( wc)h is the flux in the
moisture captured by the vegetation and 3

inversion layer. The term  ( wc)h  h  describes the layer

z
evapotraspiration. This parameterization accounts for  

seasonal evolution of the vegetation cover. It also uses
information regarding vegetation and soil types.
C. Parameterization of processes in the Planetary
Boundary Layer (PBL)
(1) PBL scheme according to Hong et al. [22]. This
scheme represents the so-called non-local diffusion
class of schemes. This scheme is described in more
detail below, in the section describing the calculation
of turbulence factors, which are used in the transport
model.
(2) Mellor-Yamada-Janic (MYJ) PBL model. This
of entrainment on the upper border of Atmospheric
Border Layer (ABL), which is assumed to be
proportional to the floatation flux in the surface layer.
The height of the ABL is defined as the height, where a
minimum flux in the inversion layer exists.
The coefficient of turbulent momentum exchange in
the ABL is written as Eq. (2):
Kzm = κwsz(1 - z/h)p (2)
where р = 2 is a constant which determines the profile, κ
= 0.4 is Von Karman constant.
The scale of velocity ws is given by Eq. (3):

ws  u*3   mw*3 z / h 
parameterization of turbulence in the boundary layer 1/ 3
(3)
and in free atmosphere is described in Refs. [17, 18]
and represents a modification of turbulence closure at where u* is surface friction velocity, and  m = 8 is a
level 2.5 based on the equation of turbulent kinetic constant.
energy balance. The turbulent momentum exchange coefficient in free
atmosphere is correspondingly defined as Eq. (4):
3. Method of Turbulence Characteristics
Calculation  U 
Km,t  l 2 f m,t (Rig)   K0 (4)
 z 
Vertical and horizontal turbulence coefficients,
The gradient Richardson number is calculated in a
which are required for the operation of the impurity
cloudless layer as in Eq. (5):
transport model, were calculated using the
g   v / z 
parameterization described in Ref. [22]. Let us Rig 
 v  U / z 2 
(5)
describe it in more detail.
and in a cloudy layer:
3.1 Calculation of Vertical Turbulence Coefficients
 L q  g2 1 ( A  B) 
Rig c  1  v v   Rig   (6)
Vertical turbulence coefficient is calculated using Rd T   U / z C pT (1  A) 
2

parameterization of the boundary atmosphere layer.
where A  Lv qv / C p Rv T and B  Lv qv / Rd T .
2 2
The closure scheme in this parameterization is based
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
Length scale:
1 1 1 represented by a large number of studied particles. The
 
l z 0 , λ0 = 30 M
For stable conditions (Rig > 0)
1 superposition of a stochastic component, which is
f t ( Rig ) 
(1  5Rig ) 2
and
Pr  1  2.1Rig
For unstable conditions (Rig ≤ 0)
8 Rig
f m ( Rig )  1 
(1  1.746  Rig )
3.2 Calculation of Horizontal Turbulence Coefficients
Calculation of horizontal turbulence coefficients is
performed according to the Smagorinsky method [23],
which considers them to be proportional to horizontal
deformation using first-order closure according to the
Eq. (11):
1/ 2
K h  Cs2l 2 0.25D11  D22   D122 
xy
2
 
where D11, D22 and D12 are components of the
deformation tensor, l  xy 
1/ 2
is length scale, Cs
is a constant (Cs = 0.25). For scalar transport the
calculated value of Kh for momentum is divided by the
turbulent analogue of the Prandtl number.
4. Description of the Transport Model
Currently there exists a number of approaches to
modeling atmospheric transport. They differ by the
factors taken into account, which affect the processes
of atmospheric dispersion, and by the ways these
factors are being accounted for.
This section contains a short description of the
transport model. A more detailed description is given
in Ref. [2].
The atmospheric dispersion model incorporated into
the NOSTRADAMUS code is a 3-dimensional
stochastic (Monte-Carlo) model of Lagrangian
trajectories.
47
In the models of this type the gas cloud is
(7) motion of each particle includes an ordered transport
by the wind allowed in the model, with the
(8) determined by turbulence. Average motion is assigned
to each particle in such a way, that the total mass (or
(9) activity) of all particles would be equal to the full mass
(or full activity) of the source. A stochastic model like
this would lose its representativeness if the number of
(10) particles is insufficient. For example, for a total number
of about 5,000-10,000 particles the surface
concentration would demonstrate appreciable
fluctuation, because the number of particles in the
volume being analyzed would be insufficient.
To obtain smoother results a special method of
smoothing of the calculated concentrations field over
space and time, based on the diffuse cloud model, was
introduced. This modification represents each test
particle as a continuous cloud with particle
(11) concentration distributed according to the Gaussian
law. In this approach concentration is calculated as a
sum of contributions of all clouds at a given point.
The shift of the cloud center during the time step is
described by the Eq. (12):
dX = U dt +(2K’y )1/2 ax, dY = V dt +(2K’y )1/2 ay,
dZ = W’ dt +(2K’z)1/2 az (12)
where ax, ay, az are random values with normal
distribution and dispersion <ax2> = <ay2> = <az2> =
dt.
Evolution of clouds (with horizontal and vertical
radii R and H) is described by the Eq. (13):
d/dt (R2) = 2 K”y, d/dt (H2) = 2 K”z (13)
where K’y, K’z, K”y, K”z are diffusion coefficients:
K’y,+ K”y = Ky, K’z + Kz” = Kz (14)
or
K’y = Ky , K”y = Ky (1  ), (15a)
K’z = Kz , K”z = Kz (1  ) (15b)
 equals to 0.9.
4.1 Calculation of Tracer Concentration
The concentration of a certain cloud with its
center coordinates xi, yi, zi at a certain point x, y, z is
48 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion

Ai   x - x i 2  y - y i 2  been performed in Karlsruhe, Germany: one in April

Ci  exp - - 
2 1.5 R 2 H  2R2 2R2  1983, another one in March 1983 and two more in
  (16)
March 1985. The exact dates and characteristics of the
   z - z i 2    z  z i 2  
*  exp  - 
  exp -

 experiments are given in Table 1.
  2H
2
  2 H 2  
In all the experiments the gas source (SF6) was
where
placed on at tower the height of 100-150 m. Gas was
t
turned on during eight time intervals of 30 minutes
R2   2 K 1    dt
0
x
each.
Gas concentration in the atmosphere was measured
is the horizontal radius of the cloud.
t
in multiple locations (Fig. 1b). Transportable gaging
  2 K 1    dt stations were located at these sites. They measured
2
H z
0
surface tracer gas concentration averaged over 30
is the vertical radius of the cloud.
minute time intervals for 3 to 5 hours in each
Ai is overall activity of the cloud.
experiment.
The sum of exponentials in the right part represents
The gaging stations were located along three arcs
“reflected particles”. They are used for correcting
(Fig. 1b) with the radii of 10-12 km, 20-23 km and
surface boundary conditions and providing mass
43-49 km (in the third experiment the second and the
conservation (without settling).
third arcs had the radii of 30-33 km and 52-69 km). The
Every particle (with a given gravitational settling
overall number of stations used in these experiments
rate) has a corresponding group of radio nuclides.
varied from 20 to 50.
The nuclide composition of each particle changes
Air samples were collected into plastic bags during
with time due to radioactive transformations. The
30 minutes and then sent to the laboratory for chemical
model accounts for decay chains of a 100 of the main
analysis. Thus it may be considered that the observed
radio nuclides. Concentration of radio nuclides in the
quantities represent surface tracer concentration
exhaust cloud may change due to precipitation
averaged over a 30 minute time period. Below these
scavenging. The present version of the code assumes
quantities are compared to concentrations, calculated
that the bottom edge of the clouds from which the
using the set of models.
precipitation occurs is located higher than the upper
As said before, measurement of meteorological
boarder of the jet of radioactive impurities, thus all the
parameters was performed on a 200 meter high tower,
jet is subject to scavenging.
located near the source tower.
Scavenging is accounted for by means of reducing
These measurements provided vertical temperature
the activity of each model particle, keeping the total
and wind profiles in the boundary layer of the
number of particles in this process unchanged.
Table 1 Characteristics of field experiments.
5. Description of Experiments Experiment No. 1 2 3 4
Date 27.04.83 23.03.84 23.03.85 25.03.85
This section provides a short description of field Tracer type SF6 SF6 SF6 SF6
experiments, performed in Germany in 1983-1985, as Source 100 100 140 140
height (m)
well as of numerical experiments performed using the
Emission 22.6 22.3 28.12 28.12
models for testing and developing the set of models. rate(g/s)
Duration 3.5 3 4.55 4.55
5.1 Description of Field Experiments (hours)
Number of 21 36 48 50
gaging stations
A series of four mesoscale field experiments has
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
atmosphere reaching the height of 200 m during the
experiments. Measurements on the tower were carried
out every 10 minutes. Wind characteristics were
recorded at the altitude of 40, 100 and 160 m.
5.2 Description of Numerical Experiments
Belikov et al. [2] provided the results of testing the
Lagrangian model of radionuclide transport described
above. In these experiments meteorological parameters
were set to the same values over space and were equal
to the values measured on the mast.
The main distinctive feature of the testing described
in this paper is the employment of non-uniform fields
of meteorological elements with the account for actual
terrain relief, corresponding to the spatial resolution of
the model.
The goal of hydrodynamic model testing was to
obtain prognostic meteorological fields which would
provide the maximal approximation to wind
observations and to evaluate the proximity between
calculated and measured concentrations. The quality of
reproduction of gas concentrations in the atmosphere
was analyzed using the combination of models
described above.
Testing of the set of models engaged the following
steps: calculation of numerical forecasts with different
versions of the hydrodynamic model, selection of
appropriate fields among the forecast results,
transformation of those fields into a coordinate frame
for the transport model and calculation of turbulence
coefficients, followed by calculation of atmosphere
tracer concentrations using the transport model. Next
an evaluation of meteorological fields and surface
concentrations was performed.
There are no fields of meteorological elements with
high spatial resolution available in Russia for the
period of field experiments (i.e. 1983-1985). At that
time computer performance capabilities in Russia were
not sufficient for performing objective analysis with
high spatial resolution and high quality. The data of
reanalyses available for these years also demonstrate
49
quite a low resolution. Nevertheless, as mentioned
above, there are results available of three reanalyses,
performed by different centers around the world: the
NCAR\NCEP, USA realnalysis, the ECWF reanalysis
by the European Center of Medium-range Weather
Forecasting and the JRA reanalysis by the
Meteorological Office of Japan. The spatial resolution
of the first two reanalyses data was 2.5 (i.e. about 250
km), and 1.25 (about 125 km) for the last reanalysis.
Preliminary experiments were performed to evaluate
the time required adapt the model to initial data.
Evaluation of surface air temperature on the mast was
used as adaptation quality criterion. It has been
demonstrated that the estimations improve after 12 to
18 hours of forecasting. In this connection the 20th
hour of forecasting was chosen as the beginning of the
field experiment and the forecast results were analyzed
for the hours 20 through 24 in all experiments in order
to minimize the spinup effect.
Since the initial data resolution was so low and the
field experiments were carried out in such a short
interval of time (only 3-4 hours), even a small local
error in the initial data in the magnitude and direction
of the wind distinctly affected the results of the
experiments. Therefore, the first series of experiments
was focused on gathering optimal initial and boundary
conditions.
Another series of experiments was focused on
gathering the optimal set of parameterizations of
physical processes in the surface and boundary
atmospheric layers, which would allow obtaining
surface gas concentrations closest to the ones observed.
6. Results of Calculations Analysis
When analyzing the experiments, the calculated
values are compared with each other and with the
values, obtained by a model of radionuclide transport
based on actual meteorological measurements from the
tower. This calculation was considered as the potential
of the radionuclide transport model. It was used in the
evaluation of experiments with combined models for
50 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
analyzing the approximation degree of the results
obtained using prognostic values compared to
calculation of transport based on observation data.
6.1 Evaluation of Prognostic Wind
Prior to comparing surface tracer concentrations
obtained in different experiments we shall discuss the
evaluation of the vertical profile of prognostic wind in
the model at the point with the coordinates of the mast.
The quality of reproduction of the vertical wind profile
in the surface and boundary atmosphere layers will, to a
great extent, define the quality of results of surface
concentrations distribution. Unfortunately, it is
impossible to evaluate the reproduction quality of
horizontal fields of the wind, because the wind was
measured only near the source mast and data from
standard synoptic or aerologic stations in the
prognostic area is unavailable for this period of
time.
Fig. 2 demonstrates a vertical distribution of mean
absolute errors averaged over the time intervals of the
field experiments (about 3 hours) in the model forecast
of wind magnitude and direction in different groups of
experiments.
The set of curves at the top represents the errors in
prognostic wind in four field experiments as a function
of initial data. All the curves were calculated for the
following combination: first method of surface layer
parameterization, second method of soil
parameterization and first method of boundary layer
parameterization. One can see that the errors in wind
speed magnitude vary from 1.5 to 5 m/sec.
One can also see that for the first field experiment
(broken lines with rectangles) the least errors in wind
magnitude were obtained in the reanalysis JRA with a
spatial resolution of 1.25 degrees. The errors decreased
with altitude.
It is significant for the second experiment (broken
lines with triangles) that in the lower troposphere
(below 140 m) the least errors were obtained in the
calculations with NCEP reanalysis, and in the higher
troposphere in the JRA reanalyses.
The results of the third experiment (broken lines
with circles) were almost the same: in the lower part of
the atmosphere (below 100 m) the least errors were
obtained in the NCEP reanalysis, and in the higher part
in the JRA reanalysis.
The fourth experiment (broken lines with diamonds)
was characterized by the largest errors in wind
magnitude. The least errors were obtained using initial
data of the ECMWF reanalysis.
Thus judging by the results of four experiments in
the lower part of the atmosphere the NCEP reanalysis
is most suitable as initial data for reproducing the wind
speed module.
The series of broken lines at the bottom of the figure
represents the quality of reproduction of the wind
direction depending on the altitude. One can see that
the average error lies between 10 and 45 degrees.
Obviously an error of 45 degrees can alter the results of
concentration calculations significantly.
One can see that in the first field experiment the
NCEP and JRA reanalyses have demonstrated almost
identical errors in wind direction, in the second
experiment the best description of wind direction was
obtained using the ECMWF reanalysis, in the third
experiment using the NCEP reanalysis, and in the
fourth also using the NCEP reanalysis. In the last case
the errors in wind direction were the highest.
Thus one may assume that in average the best
description of the wind in the lower troposphere was
obtained using the NCEP reanalysis, regardless of its
resolution not being the highest. Therefore, the
experiments on selecting an optimal set of
parameterizations of physical processes in the
atmosphere were performed using initial data of the
NCEP reanalysis.
Fig. 3 demonstrates similar vertical profiles of errors
in wind for hydrodynamic forecasts averaged over the
time period of the field experiments, where the NCEP
reanalysis was used as initial data. The profiles differ in
surface and boundary layer descriptions.
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model 51
and Lagrangian Model of Atmospheric Dispersion

200

180

160

140
m

120

100

80

60

40
0 1 2 3 m/s 4 5 6

200

180

160

140
m

120

100

80

60

40
0 5 10 15 20 25 30 35 40
Degrees

JRA_121 (1exp)200
1
180
16
40
0
ECMW F_121(1exp) NCEP_121(1exp)
м

12
0
80
0
6
40
0

0 1 2 3 4 5 6

м/ с

JRA_121 (2 exp) ECMW F_121(2exp) NCEP_121(2exp)

JRA_121 (3 exp) ECMW F_121(3exp) NCEP_121(3exp)

JRA_121 (4 exp) ECMW F_121(4exp) NCEP_121(4exp)

Fig. 2 Vertical distribution of errors in wind magnitude (top) and wind direction (bottom) averaged over the time intervals of
the field experiments obtained in the experiments with different initial data. Notations: ECMWF-initial data and boundary
conditions taken from the European Center of Medium-Range Weather Forecasting reanalysis, NCEP-from the USA National
Forecasting Center reanalysis-NCEP/DOE, JRA-from the Japanese Meteorological Office. The first index represents
parameterization of physical processes in the surface layer (1), the second index represents parameterization of physical
processes in the soil (2), the third index represents parameterization of physical processes in the planetary boundary layer (1).
52 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion

200

180

160

140
m

120

100

80

60

40
0 1 2 3 4 5 6
m/s

200

180

160

140

120
m

100

80

60

40
0 5 10 15 20 25 30 35 40
Degrees

NCEP_121(1exp)
2 00
1 80
1 60
1
1420
0
NCEP_222 (1exp)
м

1 00
80
6
40
0

0 1 2 3 4 5 6

м/с

NCEP_121(2exp) NCEP_222 (2 exp)

NCEP_121(3exp) NCEP_222 (3 exp)

NCEP_121(4exp) NCEP_222 (4exp)

Fig. 3 Vertical distribution of errors in wind magnitude (top) and wind direction (bottom) averaged over the time intervals of
the field experiments obtained in the experiments with different parameterization of the boundary and surface layers.
Notations: NCEP-initial data and boundary conditions taken from the USA National Forecasting Center
reanalysis-NCEP/DOE. The first index represents parameterization of physical processes in the surface layer (1 or 2), the
second index represents parameterization of physical processes in the soil (1), the third index represents parameterization of
physical processes in the planetary boundary layer (1 or 2).
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
This paper includes experiments with two sets of
physical processes, but actually a larger number of
experiments have been performed.
Experiment 121 comprised the following: surface
layer parameterization based on the similarity theory,
soil parameterization from the NOAH model presented
in Ref. [20] and PBL parameterization from the MRF
model described in detail in Ref. [22]. Let us refer to
this combination as “Set I”. The second set denoted as
“222” is an experiment using surface layer
parameterization from the ETA model, the same
parameterization for soil processes and the Mellora et
al. PBL parameterization described in Refs. [17, 18].
Let us refer to this combination as “Set II”. The graph
at the top shows that in the first field experiment the
errors in wind magnitude in the lower troposphere were
slightly less with Set II and comprised approximately 3
m/s, in the second field experiment the least errors
(around 1.5-2 m/s) were obtained using Set I, in the
third experiment the difference is particularly evident
and the least errors were obtained with the Set I; and in
the fourth experiment the best description of the
vertical profile (with differences between errors of over
1 m/sec) was obtained using Set II. In the fourth
experiment the errors themselves were the highest
comprising approximately 4 m/s.
The series of graphs at the bottom demonstrates
errors in the wind direction. One can see that the
differences are insignificant in the first experiment. For
the altitude of 60 to 140 m the errors for the Set I are
slightly smaller than for the Set II. In the second field
experiment errors in the wind direction for the Set I are
more than 10 degrees smaller, which may be of
fundamental importance in calculating transport in the
lower atmosphere. In the third experiment the errors
were slightly smaller with the Set II. In the fourth
experiment the errors for both sets were almost
identical.
Fig. 4 presents the Н500 maps of surface pressure
and surface altitude. From them one can see that the
area of field experiments 1, 3, and 4 were located at the
53
periphery of a cyclone and it was only for the second
experiment that the synoptic situation was
different—the area of the field experiment was located
in a col, but a cyclone was arriving. Thus it is difficult
to explain the differences in the results obtained using
two sets of parameterizations by different synoptic
conditions.
With some reserve we can conclude that on average
the Set I provides a better reproduction of the vertical
wind profile.
6.2 Analysis of Surface Concentration Maps
Fig. 5 presents the maps of surface concentrations
for each of the four field experiments. The column to
the left shows concentration maps obtained using
NCEP data, the middle column is based on JRA data,
and the column to the right presents the results of
idealized calculation based on the data observed from
the mast. Plume axes are also shown in the figures.
Table 2 presents estimations of the tilt angles of the
traces relative to the horizontal calculated in the
experiments with the prognostic model and with
observed wind. One can see that in all filed
experiments but the fourth one, the plume axis is closer
to the values calculated based on the observed wind
with data of the NCEP reanalysis, and this is confirmed
by the mean error in deviations. Fig. 5 together with the
Table 2 show that concentrations are closer to the
transport by the observed wind in experiments 1 and 4,
and the direction of surface transport is closer to the
observations when using NCEP as initial data, which
confirms the conclusion made in the previous section.
The maximal discrepancy with the transport based
on the observation data is seen in the calculations based
on JRA data in experiments 2 and 3.
Practically all traces of surface concentrations are
slightly broader, than with an empiric wind and a
turbulence, which was assumed for those calculations.
One can also see that the concentration trace
configuration depends considerably on the initial and
boundary data used in the experiments.
54 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion

Fig. 4 Н500 maps of surface pressure and surface altitude as of 27.04.1983 (top left), 23.03.1984 (top right), 23.03.1985
(bottom left), 25.03.1985 (bottom right). The filled square shows the area of calculations.
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model 55
and Lagrangian Model of Atmospheric Dispersion

NCEP 121 JRA 121 Observed wind

Experiment 1 (t = 2 hours after beginning)

Experiment 2 (t = 1 hour 20 min after beginning)

Experiment 3 (t = 1 hour 50 min after beginning)

Experiment 4 (t = 2 hours 40 min after beginning)

Fig. 5 Surface gas concentrations at the time of four field experiments, calculated using initial data from the WRF model (two
columns to the left) and based on observations of the wind on the mast (column to the right).
56 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion

6.3 Statistical Comparison of Measured and experiment with the initial data from ECMWF. The
Calculated Gas Concentrations error was the smallest in the experiment with ECMWF.
The value of the average statistical bias in the
This section provides a description of summarized
experiments with the initial data from ECMWF was
statistical characteristics of gas concentrations
almost the same as in the experiments with the
calculated for all experiments.
measured wind but its sign was opposite.
Average values for all field experiments data and for
Table 3 also shows that the best statistical
each model experiment the following statistical
characteristics which were the closest to an idealized
characteristics were calculated. Results are
calculation were obtained in the experiment with the
demonstrated in Tables 2 and 3.
initial data from ECMWF.
An average error is defined as an average difference
For the understanding of the difference between the
between two data sequences (model Ccali and measured calculation of surface gas concentrations with
Cobsi ): prognostic wind values and with the observed wind let
BIAS  Ccal  Cobs 
1 N

 Ccali  Cobsi .
N i 1
 (17)
us refer to Fig. 6. The figure demonstrates that at
A normalized mean-square error is calculated as medium (around 50-60 km) and distant (about 100-120
km) radii the surface concentrations obtained using
 Ccal 2
N
 Cobsi
NMSE 
1
 i 1
i

,
(18) prognostic values are very similar to those obtained
N Ccal  Cobs
based on the observed wind, but at the nearest radii
where the value of errors is significantly higher, tending to
N
1
С 
N
 Ci . overestimate the concentrations.
i 1

From Table 3 one can see that the percentage of Table 2 Tilt angles relative to the trace horizontal
calculated in experiments with a prognostic model and
cases, when the calculated concentrations are higher
observed wind.
than the observed concentrations (OEX) in all NCEP 121 JRA 121 Observed wind
experiment is less than in calculations with observed 42 36 46
wind. In the ECMWF121 experiment it was the 25 20 36
34 22 38
smallest being twice as small as in calculations with the 36 37 39
observed wind. The smallest percentage of cases when
Table 3 Average statistical characteristics for four field
the errors were overestimated less than twice (FA2)
experiments.
appeared to be in the version with the initial data from
Measured JRA121 NCEP121 ECMWF121
NCEP, and this percentage is more than twice as small OEX 0.51 0.30 0.25 0.24
as in the experiment with the observed wind. The FA2 0.22 0.13 0.10 0.17
FA5 0.41 0.26 0.25 0.35
percentage of cases, when calculated concentrations Pcor 0.45 0.07 0.03 0.12
are overestimated less than by 5 times (FA5), is the NMSE 4.9 18.4 27.2 17.8
smallest also in the version with the initial data from Bias -44 143 153 55
OEX: Percentage of cases, where the model values exceed the
NCEP and is almost twice as small as in calculations
experimental ones;
with the observed wind. The correlation coefficient FA2: Percentage of cases, where the errors of overestimation of
Pcor of observed concentrations and calculated Piersy calculated concentrations were less than by 2 times;
coefficient in all experiments with a model turned out FA5: Percentage of cases, where the errors of overestimation of
calculated concentrations were less than by 5 times;
to be lower than in idealized calculations with the
Pcor: Pearson correlation coefficient;
observed wind. Normalized mean error was 3 to 4 NMSE: Normalized mean-square error;
times higher. The highest coefficient was in the Bias: Mean error (ng/m3).
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model 57
and Lagrangian Model of Atmospheric Dispersion

Mean deviation at different distances

1300

1040

780

520

260
Bias, ng/m3

Standard
0 WRF

-260
0 2 4 6 8 10 12 14

-520
0 0 0 0 0 0 0

-780

-1040

-1300

Distance, km

Fig. 6 Average deviation of surface concentration calculated in the ECWMF121 experiment compared to observations from
different distances.

We assume that these deviations arise from the fact
that at small distances measurements of the wind on the
mast were rather accurate, but the initial data with
rather low resolution, which were used, did not agree
with the results of observations. At distances of about
50-100 km the errors in prediction of wind became
comparable to the errors resulting from the
non-uniformity of the wind. Moreover, this scale
corresponds to the scale of initial data and the
hydrodynamic model was able to return results which
were not worse, and in some experiments even better,
than those based on observation near the source.
Fig. 7 demonstrates a diagram of dispersion of
maximal surface concentrations based on the
observation data and on calculated date with the initial
and boundary conditions from the NCEP reanalysis.
One can see that regardless significant deviations of
some calculated concentrations, the correlation
coefficient is R2 = 0.4483.
7. Benchmark of Set of Hydrodynamic
Weather Forecast Model and Lagrangian
Model of Atmospheric Dispersion to
Simulate the Consequences of Accident at
Fukushima Daiichi Nuclear Power Plant
The situation around the Fukushima Daiichi station
was modeled as an example and for evaluation of the
possibility of using the WRF-ARW regional
hydrodynamic model and the radionuclide transport
model of the NOSTRODAMUS code.
The earthquake which took place on March 11, 2011
in Japan and had its epicenter in the Pacific Ocean
provoked a tsunami which, in its turn, leads to an
accident at a number of nuclear power plants in Japan.
One case involved release of radioactive agents (the
Fukushima Daiichi NPP).
Based on the analysis of the Fukushima Daiichi
accident, it was found that the main release of
radioactivity took place on March 15, and this release
led to a significant contamination of Japanese lands
outside the territory of the power plant. The
meteorological conditions around the Fukushima
Daiichi power plant were rather complex. Starting from
12 o’clock in the night on the 15 of March (local time)
the direction of the wind was to the south and
south-west, away from the sea. The meteorological
data also demonstrates that during the second half of
the day (starting from midday) the wind changes its
direction and aimed to the north-west. In addition,
intense rain had started on the 15 of March due to
the arrival of a cyclone from the south-west. This
58 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion

Maximal concentrations
4000

3500
using NOSTRADAMUS model

3000
Data from the experiments

2500

2000
R2 = 0.4483
1500

1000

500

0
0 1000 2000 3000 4000 5000 6000 7000
Measured data

Fig. 7 Diagram of dispersion of maximal concentrations.

combined with the wind directed towards dryland could evaluated based on modeling radionuclide
have been very unfavorable, as it could contribute in transportation. The calculated results were compared to
radioactive fall-out overland. The wind direction the measured doses obtained using aerial gamma
changed towards the Pacific only by 8 pm on March 15. photography and to the estimates of Cesium
The contamination of the Japanese lands caused by radionuclides fall-out (according to DOE, USA).
the Fukushima release on the 15 of March was Fig. 8 demonstrates that the results obtained by the

(a) (b)
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model 59
and Lagrangian Model of Atmospheric Dispersion

(c) (d)
Fig. 8 (a) Fallout radionuclide Cs, modeling results, Bq/m 2, (b) fallout radionuclide Cs, data DOE (USA) Bq/m 2, (c) annual
effective doze from external exposure, modeling results, mSv, (d) annual effective doze from external exposure, data DOE (USA)
mSv.

NOSTRADAMUS code, taking into account the
complex meteorological conditions, were comparable
to the data received from the aerial gamma
photography of the contaminated territories of Japan.
Thus, the combined use of the WRF-ARW regional
hydrodynamic model and the NOSTRADAMUS
radionuclide transport model made it possible to
evaluate the actual radiation environment, which was
caused by the radioactive release on the Fukushima
Daiichi nuclear power plant on the 15 of March 2011.
8. Conclusions
This paper presents results of numerical experiments
with a regional hydrodynamic model WRF. The goal of
the experiments with the hydrodynamic model was to
obtain the best reproduction of wind characteristics in
the southern region of Germany (near Karlsruhe),
where field experiments performed in 1984-1985 to
validate transport models.
As a result of numerous experiments with the data
from various reanalyses used as initial and boundary
conditions for the model it has been shown that the best
reproduction of vertical wind profile can be obtained
using the NCEP/DOE reanalysis.
It has also been demonstrated that the errors in
vertical distribution of the model wind are extremely
sensitive to the type of parameterization of the surface
and boundary layers used in the model.
Unfortunately, the statement that the combination
NCEP121 (i.e. using data of the NCEP/DOE reanalysis
as initial and boundary conditions, with
parameterization of the processes in the soil as
described in Ref. [20] and the procedure of
parameterization of the surface layer based on the
similarity theory, exchange coefficients for heat,
moisture and momentum fluxes calculated using
stability functions [15], and the first PBL scheme
described in detail in Ref. [22]) produces the best
results was not confirmed by all the experiments.
The paper demonstrates that using the WRF model
for predicting tracers and, in future, radionuclides,
allows to obtain prognostic concentrations of tracers
with rather realistic characteristics, but with a slightly
broader trace than in calculations with the observed
wind.
A broader trace may be due to overestimated
coefficients of horizontal turbulence, obtained using
the equation of Smagorinsky [23].
Analysis of statistical characteristics of all
experiments has demonstrated that at medium and
60 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
distant radii (around 50-100 km) errors in
concentration calculations based on prognostic wind
values are very close to the errors of an idealized
experiment, but at nearest radii the errors are
significantly higher, tending to overestimate the
concentration values.
To summarize, the results of testing the set of
models demonstrate that the combination of the
regional prognostic model WRF with high spatial
resolution together with the transport model from the
NOSTRODAMUS code allows to obtain a distribution
of surface gas concentrations with a sufficient
accuracy.
Moreover, the set of models demonstrates that in the
complex changeable meteorological conditions, the
calculated trace of radioactive fallout in the territory of
Japan correlates very accurately with observed data.
Acknowledgments
The authors would like to thank the Ministry of
Education, Culture, Sport, Science and Technology of
Japan (MEXT), the Department of Energy (DOE) and
the National Nuclear Security Administration (NNSA)
of the USA for the information provided.
The study was partly supported by RFBR foundation,
grants 09-05-00652-а and 10-08-00493-а, FP7-IRSES
“CLIMSEAS”.
References
[1] R.V. Arutyunyan, V.V. Belikov, V.N. Semyonov, O.S.
Sorokovikova, L.P. Starodubtseva, L.A. Fokin, et al.,
Testing experimental system for evaluation of atmospheric
transport of radioactive impurities using regional
high-resolution weather forecast, in: Proceedings of RAS,
Energetika, 2008, pp. 132-152. (in Russian)
[2] V.V. Belikov, V.M. Goloviznin, Y.V. Katyshkov, V.N.
Semyonov, L.P. Starodubtseva, O.S. Sorokovikova, et al.,
NOSTRADAMUS–A computer system for predicting
radiation environment, Verification of the model of
atmospheric transport of impurity, in: Proceedings of
IBRAE, Moscow, Nauka, 2008, pp. 41-103. (in Russian)
[3] V.S. Kosykh, N.V. Klepikova, I.V. Stogova, N.I.
Troyanova, G.N. Freimundt, V.A. Denkin, et al.,
Operational system for prediction of radiation
environment in Murom region in case of incident, in:
Proceedings of RAS Energetics, 2008, pp. 142-152. (in
Russian)
[4] K.G. Rubinstein, R.Y. Ignatov, E.V. Gritsan, I.V.
Novikova, R.V. Arutyunyan, V.N. Semyonov, et al.,
Using the regional non-hydrostatic model for obtaining
meteorological data for prediction of radionuclide spread
and deposition, Meteorologiya i Hydrologiya 8 (2009)
27-38. (in Russian)
[5] R.V. Arutunjan, V.N. Semenov, O.S. Sorokovikova, A.L.
Fokin, R.U. Ignatov, K.G. Rubinshtejn, Atmospheric
transport modeling with 3D Lagrangian dispersion codes
NOSTRADAMUS and meso-scale meteorological model
with height space resolution, compared with tracer
experiments at regional scale, in: International Conference
“Mesoscale Meteorology and Air Pollution”, Odessa,
2008, p. 81.
[6] R. Ju. Ignatov, K.G. Rubinstein, E.V. Gritsan, I.V.
Novikova, Using a regional model for calculation of
meteorological data for the purpose of radionuclide
distribution and deposition forecasts over the Kola
Peninsula region, in: International Conference and Young
Scientists School on Computational Information
Technologies for Environmental Sciences, Tomsk, Russia,
2007, p. 35.
[7] R. Ju. Ignatov, E.V. Gritsan, K.G. Rubinshtein, Results of
determination an optimum configuration of regional
model WRF for Moscow region, in: International
Conference “Mesoscale Meteorology and Air Pollution”,
Odessa, Sept. 15-17, 2008, p. 56.
[8] V.N. Semenov, R.V. Arutunjan, O.S. Sorokovikova, A.B.
Fokin, K.G. Rubinstein, R. Ju. Ignaov, et al., Mesoscale
model of atmospheric dispersion of contamination under
temperature and topography in homogeneity of landscape
and its usage in DSS for nuclear accidents
NOSTRADAMUS, in: International Conference and
Young Scientists School on Computational Information
Technologies for Environmental Sciences, Tomsk, Russia,
2007, p. 48.
[9] V.N. Semenov, O.S. Sorokovikova, A.B. Fokin,
Modeling of contaminant transport in a megapolis, in:
International Conference and Young Scientists School on
Computational Information Technologies for
Environmental Sciences, Tomsk, Russia, 2007, p. 28.
[10] K.G. Rubinshtein, R.Y. Ignatov, E.V. Gritsan, Using
meso-scale model for supplying forecast of transport
radionuclide, in: International Conference “Mesoscale
Meteorology and Air Pollution”, Odessa, Sept. 15-17,
2008, p. 80.
[11] E.V. Thomas, S. Vogt, P. Gaglione, Mesoscale
atmospheric experiment using tracer and tetroons
simultaneously at Kernforschungszentrum, KfK 4147
EUR 10907 EN, Karlsuhe, p. 112.
 Calculation of Surface Tracer Concentration by Set of Hydrodynamic Weather Forecast Model
and Lagrangian Model of Atmospheric Dispersion
[12] W. Ebisuzaki, J. Woolen, J. Potter, M. Fiorino, Overview
of ncep/doe reanalysis-2, in: Proceedings of the Second
WCRP International Conference on Reanalyses,
WCRP-109, WMO/TD-No. 985, 2000.
[13] R.J. Engelen, G.L. Sephens, Use of the ECMWF
reanalysis data in the assessment of the role of the
hydrological cycle in the tropical intraseasonal oscillation,
in: Proceedings of the Second WCRP International
Conference on Reanalyse, WCRP-109, WMO/TD-No.
985, Jan. 2000, pp. 201-204.
[14] K. Onogi., J. Tsutsui, H. Koide, M. Sakamoto, Sh.
Kobayashi, H. Hatsushika, et al, The JRA -25 reanalysis, J.
of the Meteorol. Soc. of Japan 85 (2007) 369-432.
[15] A.C.M. Beljaars, The parameterization of surface fluxes in
large-scale models under free convection, Quart. J. Roy.
Meteor. Soc. 121 (1994) 255-270.
[16] D.L. Zhang, R.A. Anthes, A high-resolution model of the
planetary boundary layer—sensitivity tests and
comparisons with SESAME-79 data, J. Appl. Meteor. 21
(1982) 1594-1609.
[17] Z.I. Janjic, The step-mountain eta coordinate model:
Further development of the convection, viscous sublayer,
61
and turbulent closure schemes, Mon. Wea. Rev. 122 (1994)
927-945.
[18] Z.I. Janjic, The step-mountain coordinate: Physical
package, Mon. Wea. Rev. 118 (1990) 1429-1443.
[19] S.S. Zilitinkevich, Non-local turbulent transport: Pollution
dispersion aspects of coherent structure of convective
flows, in: H. Power, N. Moussiopoulos, C.A. Brebbia
(Eds.), Computational Mechanics Publications,
Southampton Boston, 1995. pp. 53-60.
[20] F. Chen, J. Dudhia, Coupling an advanced
land-surface/hydrology model with the Penn State/NCAR
MM5 modeling system. Pt. I: Model implementation and
sensitivity, Mon. Wea. Rev. 129 (2001) 569-585.
[21] A. Xiu, J.E. Pleim, Development of a land surface model.
Pt. I. Application in a mesoscale meteorological model, J.
Appl. Meteorol. 40 (2001) 192-209.
[22] S.Y. Hong, H.L. Pan, Nonlocal boundary layer vertical
diffusion in a medium-range forecast model, Mon. Wea.
Rev. 124 (1996) 2322-2339.
[23] J. Smagorinsky, General circulation experiments with the
primitive equations, Part I: The basic experiment, Mon.
Wea. Rev. 91 (1963) 99-164.
Journal of Environmental Science and Engineering A 1 (2012) 62-67
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Changes in Physicochemical Properties during

Co-composting of Wastes with Different C/N Ratios:
Role of Effluents

Manel Kammoun-Rigane1, 2, Jean Charles Michel3 and Khaled Medhioub2

1. Faculty of Sciences, Department of Geology, Sfax University, Sfax 3000, Tunisia
2. UR Studies and Management of Costal and Urban Environments 99UR/11-36, Sfax University, Sfax 3000, Tunisia
3. UPSP EPHOR Physical Environment of HORticultural Plants, West Agrocampus 2 rue Le Nôtre, Angers 49045, France

Received: January 6, 2011 / Accepted: July 29, 2011 / Published: January 20, 2012.

Abstract: Composting of organic residue is an interesting alternative to recycle waste. The co-composting of agricultural waste is a
new management in Tunisia. Agricultural residues used in this study are mainly Mediterranean wastes. Obtained composts may be used
as soil amendments or as compounds in the horticultural and soilless substrates. The main objectives of this work were to investigate
the principal physicochemical properties during composting process of five mixtures of wastes, with different carbon/nitrogen (C/N)
ratios, prepared by the pile system and humidified with a mixture of confectionery and olive mill wastewater. During the composting
process the bulk temperature of piles was recorded. Composts sampling was assessed once a month during mesophilic and thermophilic
phases, twice during maturation phase (at 12 and 18 months). Physicochemical analyses of agricultural residues and composts samples
were established. The pH and electric conductivity (EC) were determined in a solution ratio 1/5 (v:v) and values varied from 8 to 9.8
and 0.67 to 4.3 respectively. The organic matter contents (OM) were determinated and levels varied from 55% to 20%. Total nitrogen
(TN) and carbon contents (TC) were measured by elementary analyses and values ranged from 0.62% to 1.2% and 35% to 10%
respectively. The C/N ratios of composts are related to the nature of used wastes.

Key words: Composting process, organic wastes, different C/N ratios, physicochemical properties.

1. Introduction used safely in agricultural or horticultural production.

Tunisia is one of the most important productive in
Currently, composting is a treatment used for
Mediterranean countries of olive oil and almond. So
recycling organic waste. During the process the
large quantities of wastes are generated every year as a
organic carbon form waste is easily decomposed by
consequence of olive mill factories and almond
autochthonous microorganisms. The C/N ratio is one of
shelling industries. These solid and liquid residues are
the important factors affecting composting process and
characterized by high organic matter contents and a
compost quality. It is considered C/N ratio at 25-30 as
substantial quantity of plant nutrients (N, P, K, Ca, Mg
the initial optimum ratio for composting. The
and Fe) [1, 2]. Co-composting of these wastes to
transformation of wastes into more stabilized has
provide adequate chemical composition, particularly
become an attractive option for agricultural residues as
C/N, has shown to be a suitable method which need
long considered contaminants for soil and ground
low technical and economical requirements [2, 3]. It is
water. Stabilized organic wastes called compost can be
a new management priority in Tunisia. During
composting, it is necessary to know the
Corresponding author: Manel Kammoun-Rigane, master,
research fields: compost, soil and growing media,
physicochemical, physical and biological parameters
environmental geology. E-mail: evolution to improve the process. Tejada and Gonzalez
manelriganekamoun@yahoo.fr.
 Changes in Physicochemical Properties during Co-composting of Wastes with
Different C/N Ratios: Role of Effluents
[4] showed that the use of olive mill factories wastes
composts as soil amendments could increase soil
fertility and crops production. Kammoun-Rigane et al.
[5] achieved the important role of almond shell and
olive wastes composts for tomato production in
reconstituted anthropic soil. Several studies have
demonstrated that composts derived from olive wastes
have been used as component in horticultural substrate
[6-8]. Urrestarazu et al. [9] used composted almond
shell as local rockwool substitute in soilless crop
culture and Sellami et al. [3] showed that the uses of
exhausted olive cake, poultry manure and sesame
shells compost, increased the potato yield significantly.
The aims of the present work, therefore, were to
investigate the principal modifications in
physicochemical properties i.e., pH, electrical
conductivity (EC), total nitrogen (TN), C, organic
matter contents (OM) and C/N ratios during the
co-composting processes of agricultural wastes
mixtures prepared by the pile system and humidified by
olive mill wastewater and confectionery wastewater.
The final products may be used as soil amendments or
as components in soilless substrates for horticultural
production and reconstituted anthropic soils.
2. Materials and Methods
Agricultural wastes used in this work are mainly
Mediterranean wastes generated by:
 Confectionery industries, i.e., almond shells (AS),
sesame bark (SB) and confectionery wastewater;
 Olive oil production industries, i.e., olive mill
wastewater (OMW), olive mill wastewater sludge
(OMWS) collected from the natural evaporation of
OMW in artificial basin from the region of Sfax (south
Table 1 Main physicochemical properties of raw materials.
pH OM%
AS 5.2 ± 0.1 97.9 ± 0.1
OC 5.8 ± 0.3 85.1 ± 3.6
OMWS 3.4 ± 0.5 62.9 ± 0.1
SB 5.6 ± 0.1 90.6 ± 3.8
PM 8.2 ± 0.3 26.0 ± 2.9
Values are given as a mean of three replicates ± standard deviation.
63
east of Tunisia) and olive husks (OC);
 Poultry-related industries, i.e., poultry manure
(PM).
The most important physicochemical properties of
raw materials were presented in Table 1.
SB is the most rich in nitrogen content and C/N ratio
in AS is the highest. The pH value was alkaline for PM
and acidic for the other wastes.
Five windrows were prepared by mixing agricultural
wastes CI, CII, CIII, CIV and CV. The mixtures were
prepared in the following proportions, on a fresh weight
basis and were watered by OMW and confectionery
wastewater mixture. Mixtures of about 10000 kg each
were composted in a trapezoidal windrow.
 Windrow CI: was composed of AS and SB at the
ratio of 75/25;
 Windrow CII: was prepared by mixing OH and
SB at the ratio of 75/25;
 Windrow CIII: represented a mixture of OMWS
and PM at the ratio of 70/30;
 Windrow CIV: is composed of OMWS, OH and
PM at the ratios of 50/20/30;
 Windrow CV: compost was prepared by mixing
OMWS, PM and AS (coarse) as a bulking agent at the
ratios of 55/35/10.
During the mesophilic and thermophilic phases, the
turnover of composts ensured the aeration and
homogenous humidification of waste mixtures.
Moisture was maintained at 55% using OMW and
confectionery wastewater. Effluent physicochemical
properties were recorded in Table 2.
The bulk temperature of piles was measured daily
using a thermometer (Bioblock, France) and composts
were sampled by mixing subsamples taken from the
TC% TN% C/N
43.8 ± 0.1 0.25 ± 0.05 175.2
39.5 ± 0.6 1.2 ± 0.12 33
42.0 ± 0.8 1.0 ± 0.1 42
45 ± 3.1 1.7 ± 0.06 26.47
10.6 ± 2.7 1.2 ± 0.11 9.3
64 Changes in Physicochemical Properties during Co-composting of Wastes with
Different C/N Ratios: Role of Effluents

Table 2 Main physicochemical properties of effluents.

OMW Confectionery effluent*
pH 5.40 ± 0.12 5.05 ± 0.35
EC (mS/cm) 14.15 ± 0.95 22.00 ± 1.70
Suspended solids (g/L) 14.00 ± 2.01 8.10 ± 2.70
Organic matter (%dm) 82.40 ± 1.54 99.57 ± 0.65
TOC (g/L) 36.00 ± 1.05 21.50 ± 1.63
Total nitrogen (g/L) 1.00 ± 0.08 0.66 ± 0.03
*Hachicha et al. [10], dm: dry matter.

top, the medium and the bottom portions of each
windrow. Compost sampling was assessed once a
month during the mesophilic and thermophilic phases,
and twice during the maturation phase (at 12 and 18
months).
Physicochemical analyses of agricultural residue
and compost samples were established using the
standard, EN Pr 13040 [11]. Water content was
determined by weight loss of samples, which were
dried at 105 °C for 48 h. The pH and electric
conductivity (EC) were determined in a solution ratio
1/5 (v:v) according to standards EN Pr 13037 [12]
and EN Pr 13038 [13], respectively. The organic matter
contents (OM) were assessed by determining the
loss-on-ignition at 450 °C over six hours. Total
nitrogen (TN) and carbon contents (TC) were
measured by elementary analyses based on the
standard, NF ISO 10694 [14] and NF ISO
13878/X31-418 [15] respectively, using a composts CIII, CIV and CV mainly composed by
Thermoquest® analyser.
3. Results and Discussion
3.1 Temperature Evolution
Temperatures of the mixtures increased at the
beginning of the process to thermophilic values (Fig. 1).
Then the temperature reached the optimal values
composting was considered finished when the
temperature of these windrows was stable and close to
that of the surrounding atmosphere [2, 18]. At this stage
we substituted the addition of wastewater by water.
3.2 Evolution of pH and EC
As shown in Table 3, composts pH were alkaline
with values ranging from 8 to 9.7 during the process. It
may be explained by the alkaline hydrolysis of K and
Na salts from wastes [19]. This finding correlates with
previous studies [20]. The EC values increased slightly
during the first four month for the windrows CI and CII
(Table 3). This increase may be related to the loss of
weight and accumulation of salts such as phosphate and
ammonium ions through the decomposition of organic
substances [19] and also salt providing from effluents
characterized by a high EC used for windrows
humidification. Inversely, decreases were observed for
OMWS and PM whilst OMW use. This fact is related
to a leaching effect which occurred while using water
Ambient T °C CI
CII CIII
80 CIV CV
60
T ( C)
T°C

(65-70 °C) during few weeks [16]. Zhang and He [17] 40

°

considered that level contributed to the elimination of

20
pathogens and harmful seeds. Indeed, the duration of
process phases was related to the proportion of readily
0
degradable carbon in substrates. The decrease in 0 40 80 120
temperature was recorded approximately after three Composting days

months of composting. The bio-oxidative phase of Fig. 1 Temperature profiles during the composting process.
 Changes in Physicochemical Properties during Co-composting of Wastes with 65
Different C/N Ratios: Role of Effluents

Table 3 Evolutions of pH and EC of composts during the process.

CI CII CIII CIV CV
CM pH EC pH EC pH EC pH EC pH EC
0 9.7 ± 0.2 0.67 ± 0.13 8.6 ± 0.1 1.03 ± 0.12 9.0 ± 0.2 4.4 ± 0.4 8 ± 0.2 3.4 ± 0.2 9.3 ± 0.1 3.2 ± 0.13
1 9.4 ± 0.1 0.67 ± 0.08 9.4 ± 0.2 0.97 ± 0.24 8.4 ± 0.4 4 ± 0.3 9.4 ± 0.2 2.4 ± 0.2 9.5 ± 0.1 2.2 ± 0.18
2 9.5 ± 0.1 0.64 ± 0.09 9.7 ± 0.0 1.18 ± 0.19 9.4 ± 0.1 3.7 ± 0.2 8.6 ± 0.0 3.4 ± 0.2 8.4 ± 0.1 3.4 ± 0.22
3 9.5 ± 0.2 0.84 ± 0.19 9.6 ± 0.1 1.7 ± 0.31 9.1 ± 0.2 4.3 ± 0.4 9.6 ± 0.1 2.6 ± 0.2 8.4 ± 0.2 1.6 ± 0.11
4 9.5 ± 0.2 0.75 ± 0.2 9.6 ± 0.1 1.2 ± 0.17 8.7 ± 0.1 3.9 ± 0.2 9.6 ± 0.1 2.3 ± 0.2 9.3 ± 0.3 1.6 ± 0.2
5 9.5 ± 0.1 1.26 ± 0.23 9.5 ± 0.1 2.9 ± 0.38 9.4 ± 0.3 4.0 ± 0.4 9.5 ± 0.2 1.8 ± 0.2 9.4 ± 0.2 1.8 ± 0.16
6 9.5 ± 0.2 1.18 ± 0.08 9.6 ± 0.2 2.05 ± 0.41 9.7 ± 0.1 2.6 ± 0.2 9.7 ± 0.2 2.6 ± 0.2 9.7 ± 0.1 1.6 ± 0.21
12 9.7 ± 0.2 1.07 ± 0.21 9.6 ± 0.1 1.89 ± 0.3 9.7 ± 0.1 2.8 ± 0.3 9.8 ± 0.0 1.6 ± 0.2 9.7 ± 0.2 1.6 ± 0.14
18 9.4 ± 0.1 1.61 ± 0.11 9.2 ± 0.3 1.66 ± 0.16 9.5 ± 0.2 2.6 ± 0.2 9.4 ± 0.2 2.9 ± 0.2 9.1 ± 0.0 3.3 ± 0.23
EC: electric conductivity, CM: composting months; Values are given as a mean of three replicates ± standard deviation.
Table 4 Evolutions of N, C and OM contents and C/N ratios during the composting process.
Evolution of N contents for CI, CII, CIII, CIV and CV
CI CII CIII CIV CV
0 0.7 ± 0.05 1.1 ± 0.08 1.1 ± 0.1 0.8 ± 0.12 1.0 ± 0.03
1 0.6 ± 0.01 0.9 ± 0.12 1.1 ± 0.0 0.9 ± 0.09 1.1 ± 0.04
2 0.62 ± 0.03 1.1 ± 0.02 1.2 ± 0.1 1.0 ± 0.05 0.9 ± 0.01
3 0.62 ± 0.04 0.9 ± 0.08 1.2 ± 0.1 0.9 ± 0.03 1.1 ± 0.07
4 0.63 ± 0.02 1.0 ± 0.05 1.2 ± 0.0 0.9 ± 0.05 1.1 ± 0.01
5 0.70 ± 0.01 1.0 ± 0.09 1.0 ± 0.0 1.0 ± 0.11 1.2 ± 0.03
6 0.71 ± 0.03 1.0 ± 0.06 1.2 ± 0.0 1.0 ± 0.08 1.1 ± 0.06
12 0.70 ± 0.01 1 ± 0.08 1.0 ± 0.1 1.0 ± 0.02 1.0 ± 0.09
18 0.64 ± 0.02 1.0 ± 0.1 0.9 ± 0.1 0.9 ± 0.06 1.0 ± 0.06
Evolution of C contents for CI, CII, CIII, CIV and CV
CI CII CIII CIV CV
0 21 ± 1 30.8 ± 0.2 14 ± 0.1 15.3 ± 0.8 14.7 ± 1.2
1 22 ± 0.9 21.5 ± 0.6 20 ± 0.4 14.0 ± 0.4 17.3 ± 0.9
2 21 ± 1.2 18.9 ± 0.4 17 ± 0.6 15.5 ± 0.4 17.5 ± 0.7
3 21 ± 0.8 18.9 ± 0.3 16 ± 0.5 14.8 ± 1.0 22.5 ± 1.2
4 15 ± 1.1 21.1 ± 0.9 18 ± 0.5 14.9 ± 0.09 19.5 ± 0.1
5 15 ± 0.7 17.4 ± 0.2 12 ± 0.1 15.0 ± 1.2 16.1 ± 1.1
6 17 ± 1.2 14.7 ± 0.2 16 ± 0.7 14.0 ± 0.7 15.6 ± 1.3
12 12 ± 0.6 13.9 ± 0.6 11 ± 0.9 11.1 ± 0.6 11.2 ± 2.1
18 14 ± 1.3 13.1 ± 0.3 10 ± 0.5 9.7 ± 0.2 11.9 ± 1.0
Evolution of OM contents for CI, CII, CIII, CIV and CV
CI CII CIII CIV CV
0 43.3 ± 1.2 55.1 ± 2.2 34.2 ± 1.2 35.0 ± 1.2 38.7 ± 2.2
1 42.3 ± 1.8 52 ± 3.5 31.8 ± 2.3 37.2 ± 1.8 36.2 ± 3.3
2 39.5 ± 1.5 50.8 ± 3.0 29.3 ± 1.0 34.6 ± 2.4 39.5 ± 1.9
3 37.8 ± 1.6 47.6 ± 1.6 30.5 ± 1.4 31.9 ± 0.2 37.3 ± 2.7
4 40.2 ± 2.2 37.8 ± 1.4 31 ± 2.2 27.5 ± 1.6 33.0 ± 1.2
5 34.0 ± 0.9 38.1 ± 1.0 29.2 ± 0.8 23.3 ± 1.3 31.3 ± 3.2
6 31.4 ± 1.2 33.4 ± 0.9 29.3 ± 0.7 22 ± 0.9 28.9 ± 1.5
12 32.9 ± 0.2 33.5 ± 1.3 24.7 ± 0.5 20.6 ± 0.4 20.6 ± 1.2
18 32.0 ± 0.2 30.3 ± 1.1 24.7 ± 1.1 20.2 ± 0.7 20.4 ± 1.3
Evolution of C/N ratios for CI, CII, CIII, CIV and CV
CI CII CIII CIV CV
0 31 28.2 14.6 18.8 14.7
1 37 22.6 19.3 14.9 15.2
2 34 17.9 14.4 15.5 19.2
3 34 19.8 13.6 16.7 20.2
4 24 20.6 15.2 18.8 18.4
5 23 16.5 12.1 16.0 13.6
6 24 14.1 13.6 14.0 14.8
12 19 13.3 12.3 11.5 11.2
18 20 12.7 11.8 12.6 11.9
66 Changes in Physicochemical Properties during Co-composting of Wastes with
Different C/N Ratios: Role of Effluents
for irrigating during the maturation phase. Although
EC values were around 1.6 mS/cm in AS, OH and SB
compost at the end of the process and were ranged from
2.5 to 3.1 mS/cm for OMWS and PM-basing composts.
3.3 Evolution of C, N, OM and C/N Ratios
The N, C and OM contents evolution during
composting process were presented in Table 4.
Physicochemical properties of the mixtures at the
process starting were significantly affected by the
nature of wastes. The OM contents in composts were
higher in AS, OH and SB composts (CI and CII) than
that in OMWS and PM-based composts, while
significant difference was observed in the C, N and
C/N ratio at the start of the process (Table 4). The
composts CI containing AS as substrate present the
lowest N content, whilst no significant difference was
observed in the TN contents during the bio-oxidative
phase for all composts. The C, OM contents and C/N
ratio showed a significant decrease throughout the
composting process for CII. However, the
bio-degradation of AS substrates which contain more
resistant compounds in CI was less intense mainly
during the first three months and it presents an increase
in CIII, CIV and CV. This fact could be related to low
C/N ratios in OMWS and PM characterized by low
quantities of carbon-rich materials. Therefore, the use
of OMW and confectionery effluent having a large
quantity of organic and easily degradable substances
increased C contents remarkably in these windrows
and improved C/N ratio considered adequately for
monitoring OM degradation. Then, decreases in C
contents and C/N ratio were observed. Through the
process, the C/N ratio showed a significant decrease
measured as more than 45% for AS, OH and SB less
than 35%, for OMWS and PM-based composts.
Therefore, the C/N ratios at the end of the process were
stabilized around 12 for olive wastes based composts
and about 20 for AS based compost. Bernal et al. [20]
suggested that a C/N ratio of 20 would reflect a
satisfactory degree of compost maturity, though,
Mustin [21], considerd that a value below 12 indicates
a high degree of compost maturity.
4. Conclusions
It can be concluded that used agricultural wastes can
be successfully co-composted and humidified by
OMW and confectionery wastewater. The N and C
contents and C/N ratios are related to the nature of the
co-composted wastes and values were significantly low
in olive wastes-based composts. The effluents addition
was necessary for improving the process mainly for
OMWS and PM mixtures by increasing C and OM
contents and C/N ratios. These values rise also by
adding AS as bulking agent in CIV. The OM
mineralization was less important in these composts
compared to AS, OH and SB-based compost due to the
low initial contents. At the end of the process C/N
ratios were stable and were ringing from 20 to 11 with
the highest values for AS based-compost, indicating
maturity of OM in the studied composts. The pH levels
were alkaline for all composts. The EC values were
higher in OMWS and PM-based composts compared to
AS, OH and SB based-composts, hence indicating a
possible toxicity for plants. This phytotoxicity could be
related to the high salts contents if we used compost as
principal components in growing media or if we
applied composts by high quantities for soil improvers.
Further works should be undertaken to optimize the
rate of adequate rate for using composts as soil
amendments and/or components in horticultural
substrates.
References
[1] M. Kammoun-Rigane, T. Mnif, K. Medhioub, Organic
agricultural wastes composts: Evaluation of
physicochemical properties and agronomic potential, in:
Proceedings of the International Congress of Solid Waste
Management & Sustainable Development, Tunisia, 2008,
pp. 121-126. (in French)
[2] A. Garcıa-Gomez, A. Roig, M.P. Bernal, Composting of
the solid fraction of olive mill-wastewater with olive
leaves: Organic matter degradation and biological activity,
Bioresource Technology 86 (2003) 59-64.
 Changes in Physicochemical Properties during Co-composting of Wastes with
Different C/N Ratios: Role of Effluents
[3] F. Sellami, R. Jarboui, S. Hachicha, K. Medhioub, E.
Ammar, Co-composting of oil exhausted olive-cake,
poultry manure and industrial residues of agro-food
activity for soil amendment, Bioresource Technology 99
(2008) 1177-1188.
[4] M. Tejada, J.L. Gonzalez, Effects of foliar application of a
byproduct of the two-step olive oil mill process on maize
yield, Agronomie 23 (2003) 617-623.
[5] M. Kammoun-Rigane, K. Medhioub, T. Mnif, Exogenous
organic matter: Short-term effects on amended soil
irrigated by the drips system in pilot scale, in: Proceedings
of the 10th International Meeting on Soils with
Mediterranean Type of Climate, Liban, 2009, pp.
112-115.
[6] M. Papafotiou, M. Phsyhalou, G. Kargas, I. Chatzipavlidis,
J. Chronopoulos, Olive-mill wastes compost as growing
medium component for the production of poinsettia,
Scientia Horticulturae 102 (2004) 167-175.
[7] F. Herrera, J.E. Castillo, A.F. Chica, L. López Bellido, Use
of municipal solid waste compost (MSWC) as a growing
medium in the nursery production of tomato plants,
Bioresource Technology 99 (2008) 287-296.
[8] M. Kammoun-Rigane, J.C. Michel, P. Morel, K.
Medhioub, Properties of Tunisian wastes composts
assessment for use as components in soilless substrate, in:
Proceedings of the Soil Fertility and Crop Nutrition
Management in Mediterranean Organic Agriculture,
Tunisia, 2010, p. 69.
[9] M. Urrestarazu, G.A. Martínez, M.C. Salas, Almond shell
waste: Possible local rockwool substitute in soilless crop
culture, Scientia Horticulturae 103 (2005) 453-460.
[10] S. Hachicha, F. Sellami, J. Cegarra, R. Hachicha, N. Drira,
K. Medhioub, et al., Biological activity during
co-composting of sludge issued from the OMW
evaporation ponds with poultry manure physico-chemical
characterization of the processed organic matter, Journal
of Hazardous Materials 162 (2009) 402-409.
[11] Sample preparation for chemical and physical tests,
67
determination of dry matter content, moisture content and
laboratory compacted bulk density, in: Soil Improvers and
Growing Media, European Committee for Standardization,
European Standard 13040, Brussels, 1999.
[12] Determination of pH, in: Soil Improvers and Growing
Media, European Committee for Standardization,
European Standard 13037, Brussels, 1999.
[13] Determination of electrical conductivity, in: Soil
Improvers and Growing Media, European Committee for
Standardization, European Standard 13038, Brussels,
1999.
[14] Quality of soil—Determination of carbon contents with
dry combustion (“elementary analysis”), Theme:
Chemical properties of soils, NF ISO 10694/NF X31-409,
AFNOR, 1995. (in French)
[15] Quality of soil—Determination of nitrogen contents with
dry combustion (“elementary analysis”), Theme:
Chemical properties of soils, NF ISO 13878/X31-418,
AFNOR, 1995. (in French)
[16] C. Paredes, J. Cegarra, M.P. Bernal, A. Roig, Influence of
olive mill wastewater in composting and impact of the
compost on a Swiss chard crop and soil properties,
Environment International 31 (2004) 305-312.
[17] Y. Zhang, Y. He, Co-composting solid swine manure with
pine sawdust as organic substrate, Bioresource
Technology 97 (2006) 2024-2031.
[18] S. Hachicha, M. Chtourou, K. Medhioub, E. Ammar,
Compost of poultry manure and olive mill wastes as an
alternative fertilizer, Agronomie 26 (2006) 135-142.
[19] N. Abid, S. Sayadi, Detrimental effects of olive mill
wastewater on the composting process of agricultural
wastes, Waste Management 26 (2006) 1099-1107.
[20] M.P. Bernal, C. Paredes, M.A. Sanchez-Monedero, J.
Cegarra, Maturity and stability parameters of compost
prepared with a wide range of organic wastes, Bioresource
Technology 63 (1998) 91-99.
[21] M. Mustin, Le Compost: Gestion de la Matière Organique,
François Du Busc., Paris, 1987, p. 954. (in French)
Journal of Environmental Science and Engineering A 1 (2012) 68-77
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Quantitative Screening and Resolution of Carbamic and

Organoposphate Pesticides Mixture in Extra Virgin Olive
Oil by Acetylcholinesterase-Choline Oxidase Sensor

D. Albanese1, M. Di Matteo1 and R. Pilloton2

1. Dipartimento di Ingegneria Industriale, Università di Salerno, Via Ponte Don Melillo, Fisciano, (SA) 84084, Italy
2. ENEA, C.R. Casaccia, Via Anguillarese 301, I-00123, Roma, Italy

Received: March 28, 2011 / Accepted: May 19, 2011 / Published: January 20, 2012.

Abstract: Acetylcholinesterase-choline oxidase amperometric bi-enzymatic biosensor in a flow injection configuration has been
assembled in order to quantitatively detect carbamic and organoposphourus pesticides mixture in extra virgin olive oil samples. The
recognition of the two different pesticide classes by this biosensor has been carried out by exploiting the well-known different
inhibition mechanism of carbamic (reversible inhibition) and organoposphourus (irreversible inhibition) compounds versus
acetylcholinesterase activity. In order to eliminate extraction clean-up steps for pesticides detection in extra virgin olive oil, inhibition
calibration curves and analyses of spiked real samples were performed using hexan as the carrier in the flow injection apparatus. The
results of the calibration curves showed a different inhibition power versus acetyl cholinesterase in function of the specific compound
and found that a probable interaction exists between two or more compounds that affected the toxic power of the single pesticide. The
limit of detection (5 μg·L-1) obtained for the tested organoposphourus (Paraoxon; Fenitrothion) and carbamic (Methomyl; Carbaryl)
pesticides was lower than residues established by European legislation, thus suggesting that the bi-enzymatic biosensor, extensively
described in literature, can be useful and economically convenient for quantitative screening methods in food security monitoring.

Key words: Food analysis, extra virgin olive oil, pesticides, biosensor.

1. Introduction them being destroyed as a consequence of disease and

infestation. A possible consequence of their use may be
Extra virgin olive oil (EV) is one of the most known
the presence of pesticide residues in the EV extracted
and typical food products of the Mediterranean regions
from treated fruits. Maximum residue levels have been
and Spain, Italy and Greece represent the main EV
established by different institution (Food and Drug
producers in the world. Its popularity depends on its
Administration, Codex Alimentarius Commission) in
appreciated sensorial properties besides positive
order to avoid possible risks for human health (ECs No.
effects for human health. EV total quality depends on
396/2005; No. 178/2006 and No. 149/2008) [1-3]. A
multiple factors such as olive cultivars, production
recent research [4] about pesticides contamination in
process, preservation methods and, last but not least,
commercial Italian EV olive oils shows 1.9% of
agronomic practices. Organoposphourus (OP) and
irregular samples, 3.4% of regular samples with only
Carbamic (CB) pesticides are normally used in olive
one class of residues and 12% of regular olive oil
fruit production to keep crops healthy and to prevent
samples with more than one residue, thus confirming
the importance of the development of multi-residue
Corresponding author: M. Di Matteo, full professor, main analytical methods for the determination of pesticides
research fields: shelf-life extension of food products,
optimization and process innovation for the food industry, in this important food product.
development of sensors for monitoring and control of food The most frequently used methods for detection of
quality. E-mail: mdimatteo@unisa.it.
 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
OP and CB pesticides are based on gas and liquid
chromatography in combination with mass
spectrometry. While being sensitive and reliable, such
methods require expensive laboratory equipments,
trained personnel and complex procedures to treat and
extract the real matrices.
The use of biosensors as an analytical method to
determine the presence of neurotoxic pesticides
represents an attractive alternative for research because
of their high sensitivity, selectiveness, ease and
rapidity of use. These analytical devices, based on the
intimate contact between a biorecognition element that
interacts with the analyte of interest and a transducer
element that converts the biorecognition event into a
measurable signal, are gaining momentum as
complimentary screening assays together with classical
analytical techniques due to their high selectivity and
sensitivity, low instrumentation cost, easier procedures
and rapidity of the assay. After the rapid screening step
performed with biosensors on many real samples, only
the positive ones can be performed with the classical
analytical techniques with the aim of lowering the cost
of field analysis. Among the different types of
biosensors, the electrochemical ones are especially
interesting due to the high sensitivity inherent to the
electrochemical detection and the possibility to
miniaturize the required instrumentation, providing
compact and portable analysis devices.
Electrochemical, mono or bi-enzyme, biosensors able
to recognize CBs and OPs by specific inhibition of
acetyl cholinesterase (AChE) as biorecognition
element have been extensively reported [5-10]. In both
categories OPs and CBs determination is performed by
AChE activity measurements before and after exposure
to a sample (incubation) and calculation of the
inhibition due to OP and CB compounds. Performance
of bi-enzymatic biosensor was based on two enzymatic
reactions in series as reported in Eqs. (1) and (2).
AChE
Acethylcholine + H2O acetate + choline (1)
ChOx
Choline + 2O2 betaine aldehyde + 2 H2O2 (2)
69
OP and CB lower the enzyme activity of the first
enzyme and therefore the production of choline and
subsequently reduce the current produced by oxidation
of hydrogen peroxide. In addition the enzyme (AChE)
used in this biosensor is the same which determines
neurotoxic effects in humans and animals and, for this
reason, the measurements are directly related to the
biological effect while the traditional chromatographic
methods detect the concentration of the pesticides
without any information about the neurotoxicity due to
the mix of different compounds. For inhibition
biosensors, the calibration curves and the
measurements on real samples are conventionally
expressed in concentration units of a pesticide which
gives the same neurotoxic effect.
Research and development on biosensors for
pesticides had many improvements in the last two
decades and are still going for ever increasing
complexity of the immobilization strategies of the
enzymes [11], the use of novel biological mediators as
well as bacteria [12] and miniaturization of the
transducers [13]. Bachmann et al. [14, 15] proposed
disposable screen-printed amperometric multielectrode
biosensors based on inhibition of different types of
native and recombinant AChEs and chemometric data
analysis using Artificial Neural Networks (ANNs).
Nevertheless these advanced studies of the applications
to food industry requirements and real samples remain
poor and not much explored, especially for screening of
EV production. Moreover, the analysis of pesticide
residues in EV is complicated, because of the inherent
complexity of the matrix, mainly comprising
triglycerides (98%-99%). Many pesticides are
fat-soluble non-polar compounds and tend to
concentrate and remain in the fat matrix. This behavior
normally requires the use of additional extraction
clean-up steps, to isolate or to extract the pesticide
fraction from the whole fatty matrix, prior to
conventional [16] or biosensor analysis [17]. The main
strategies used for the isolation of pesticides in olive oil
samples involve the application of techniques called
70 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
liquid-phase extraction (LPE) or liquid-solid extraction
(LSE). In both cases the use of organic solvents such as
hexan, acetonitrile was required [16]. These remarks
highlight the advantage to develop pesticide biosensor
working in organic solvent able to analyze food samples
with relatively high fat content (i.e. > 15%), such as
olive oil. Previous researches [8, 18-20] have shown the
capability of acetylcholinesterase-choline oxidase
bi-enzymatic biosensor working in non-aqueous
solvents to obtain small or no inhibition effect
concerning acetylcholinesterase activity in different
organic solvents.
An important goal which could be obtained with
amperometric biosensors based on inhibition of AChE
is the capability to distinguish between pesticide
classes (OP and CB) when both are present in a mixture.
The aim of this research, where we used a basically
assembled biosensor as extensively reported in
literature, has been mainly focused on the development
of a simple, inexpensive, and sensitive screening
method based on bi-enzymatic (AChE-ChOx)
amperometric biosensor able to obtain quantitative
information with regard to the simultaneous presence
of OPs and CBs in olive oil samples. The recognition of
the two different pesticide classes by AChE-ChOx
biosensor has been carried out by exploiting the
well-known different inhibition mechanism of CB and
OP compounds versus AChE activity. CB compounds
form an unstable bond with esterasic site of AChE that
spontaneously tends to break down in an interval time
ranging from few minutes to one hour (reversible
inhibition). OP compounds, on the contrary, form a
highly stable covalent bond phosphate-enzyme,
bringing about an irreversible inhibition [21].
Inhibition tests on AChE were carried out with OPs
and CBs solutions at different concentration in hexan,
whereas the tests on real samples were straight
followed on spiked EV sample diluted with hexan
solvent without no extraction step. To improve the
speed and reproducibility of the analysis, the
bi-enzymatic biosensor was connected to a Flow
Injection Analysis (FIA) manifold.
2. Experiment
2.1 Reagents
Acetylcholinesterase (E.C. 3.1.1.7) Type V-S from
electric eel (1000 units·mg-1 protein), Choline oxidase
(E.C. 1.1.3.17) from Alcaligenes species (10 units·mg-1
solid), acetylcholine chloride, choline chloride,
glutaraldehyde (50% aqueous solution),
pyridine-2-aldoxime methiodide (PAM),
D-aminopropyl CPG (200-400 mesh, average pore size
of 500 Å, average amine content of 70 μmol·g-1)
Bovine Serum Albumin (BSA), Paraoxon, Fenitrothion,
Methomyl, Carbaryl, were purchased from Sigma (St.
Louis, USA). Hexane for the preparation of pesticide
solutions and for the dilution of olive oil samples were
HPLC grade. All other chemicals were of analytical
grade or better. Residue-free virgin olive oil, certified
by an accredited laboratory for the absence of pesticide
residues, was provided by a local company.
2.2 Screen Printed Electrodes
Screen printed electrodes were produced in three
steps, by screen printing different consecutive layers on
transparent polyester films (Kemafoil® Coveme, Italy).
A first layer of a graphite ink was deposed to define the
conducting track and the working electrode, the second
one was a silver/silver chloride ink used as reference
electrode, while the third layer consisted in an
insulating ink, UV polymerizable, resistant to organic
solvents. Between the first and the second screen
printing step, the strips were cured at 80 °C for 25 min.
The insulating ink was polymerized by UV light (mod
HB 311 Philips, Netherlands).
2.3 Preparation of Bi-enzymatic Biosensor
Well known procedures from the literature based on
glutaraldehyde were used for immobilization of both
enzymes on the electrode surface and on microreactor
[22]. AChE immobilization was made on CPG. The
linking ability of this reagent with two opposite
 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
aldehydic groups was used to create alkyl bridges
between the amino group of the glass surface and
AChE. These bridges can be obtained with a two steps
procedure and using a high concentration of
glutaraldehyde.
300 µL of 2% solution of glutaraldehyde in
phosphate buffer (PB) 0.1 M, pH 6.8 was added to a 65
mg of CPG. The mixture was shaken for 1 h and the
activated glasses were rinsed with PB. Next, specific
amounts of AChE in PB (24, 12, 6 and 3 units) were
added to activated CPG glasses and the mixture was
stored at 4 °C overnight. AChE immobilized was then
inserted into a borosilicate glass column (L × I.D. 25
mm × 3 mm, Omnifit England) and connected to the
FIA manifold.
The crosslinking ability of glutaraldehyde was used
in a one step procedure where the enzyme (ChOx) is
mixed with BSA and a lower concentration of the
crosslinking agent to prevent its denaturation. ChOx
immobilization was carried out directly on screen
printed electrode by dropping on the working electrode
5 µL of a mixture of BSA (0.2 µg·µL-1):
Glutaraldehyde (2.5%) and ChOx (0.2 µg·µL-1)
(20:10:5). SPE was stored at room temperature for 1 h
and successively at 4 °C overnight.
2.4 FIA System
FIA configuration used is shown in Fig. 1. The
carrier stream was pumped through the FIA system by
a peristaltic pump with four channels (Miniplus 3,
Gilson, France), two of which were used for the fluxing
PB and organic solvent pesticide solutions separately.
Acetylcholinchloride solution was injected into
bi-enzymatic biosensor by manual sample injection
valve (six port loop inject valve, Omnifit, England) at
0.5 mL·min-1. Acetylcholinchloride solution was
injected into bi-enzymatic biosensor by manual sample
injection valve (six port loop inject valve, Omnifit,
England) at 0.5 mL·min-1, and the current
corresponding to the oxidation of hydrogen peroxide
was measured by a potentiostat Palm Sens (Palm
71
Fig. 1 Schematic diagram of bi-enzymatic (AChE-ChOx)
flow injection biosensor: a-PB; b-pesticide solution;
P-peristaltic pump; I-injection valve; AChE-R, AChE
reactor; V3-three way valve; PW-pesticide waste;
EC-electrochemical cell (ChOx immobilized on screen
printed electrode); W-waste; Rec recorder.
Instruments BV, Netherland).
A three way valve was inserted after AChE reactor
and before the biosensor, in order to avoid contact
between pesticides solution and ChOx immobilized on
SPE. The components of the flow injection system
were connected with 1.6 mm OD × 0.3 mm ID
Polytetrafluoroethylene (PTFE) tubing.
2.5 Determination of OP and CB Pesticides by
Inhibition Biosensor
The recognition and quantification of OP and CB
compounds by bi-enzymatic biosensor is based on the
inhibitory power of both pesticide classes versus the
activity of acetylcholinesterase [23]. On the basis of
two enzymatic (Eqs. (1) and (2)) in series it is possible
to assay the AChE activity by hydrogen peroxide
amount that is oxidized on the electrode surface at 0.7
V vs Ag/AgCl pseudo-reference electrode.
The inhibitory action of CB and OP compounds on
AChE activity is performed by a minor choline amount
resulting in a consequent reduction of the current
produced by hydrogen peroxide. The recognition of the
two different pesticide classes by ChE-ChOx biosensor
has been carried out by exploiting the different
inhibition mechanism of CB and OP compounds versus
AChE activity. Previous research on the inhibition of
AChE sensors by pesticides reported that incubation
times between the immobilized enzyme and pesticides
ranging from 10 min [8, 17] and 30 min [24, 25]. In our
experiments inhibition tests were carried out by
flowing at 0.1 mL·min-1 for 45 min (flow and time
experimentally optimized, data not shown), in FIA
72 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
manifold, hexane solutions of Paraoxon, Fenitrothion,
Methomyl, Carbaryl, alone and in mixture (Paraoxon +
Methomyl; Fenitrothion, + Carbaryl). The influence of
pure hexane on the enzymatic activity of immobilized
AChE was tested.
After 45 min of incubation in pure no significance
variations in sensor response were detected if
compared with the initial value. These data are in
accordance with previous research [8]. Through a three
way valve (V3, Fig. 1), positioned after AChE reactor,
the pesticide solutions were discarded prior to
converging on ChOx biosensor. The FIA manifold was
then rinsed for 2 minutes with PB. The residue activity
of AChE was measured by injection of Acetylcholine
chloride (AChCl) 0.5 mM and on the basis of two
enzymatic Eqs. (1) and (2) in series the heights of the
amperometric peaks were collected. In the case of
solutions with a single concentration of CB or OP
compounds the percentage of inhibition I% was
expressed by Eq. (3):
i  if (3)
I%  0  100
i0
where i0 and if represent the height of the peak
corresponding to oxidization of hydrogen peroxide
before and after CB or OP pesticides respectively.
When a mix of both pesticide classes is introduced in
flow system, the peak decrease depends on the
inhibitory effect of CB and OP alike. The following
procedure was then used to calculate the AChE
inhibition of both pesticides classes:
 Injection of Acetylcholine 0.5 mM and measure
of the biosensor response (i0);
 Introduction of CB and OP in mix at given
concentration;
 Waiting for 45 minutes;
 Injection of Acetylcholine 0.5 mM and measure of
the biosensor response (ib);
 Waiting for 30 minutes: this time was calculated
on the basis of experimental tests for the recovery of
96.09% ± 3.44% of amperometric signal
corresponding to Acetylcholine 0.5 mM before the
inhibition with methomyl and carbaryl in range of 5-20
µg·L-1;
 Injection of Acetylcholine 0.5 mM and measure of
the biosensor response (ia).
The percentage of inhibition I% of CB and OP was
expressed by Eqs. (4) and (5) respectively:
i i (4)
I % CB  b a  100
i0
i i (5)
I % OP  0 b  100
i0
where i0 represents the biosensor response in absences
of pesticides, ia is the biosensor response after the
interaction between AChE and the pesticides in mix, ib
is the biosensor response after the recovery of the
AChE inhibition caused by OP compounds. A typical
biosensor response obtained through the procedure
described above was reported in Fig. 2.
2.6 Reactivation of Inhibited AChE
The reactivation ranging from 98.5% and 100% of
inhibited AChE after contact with solutions at different
amounts of OP pesticide was carried out flowing PAM
0.1 M at 0.1 mL·min-1 for 45 min followed by the PB at
0.5 mL·min-1 for 20 min into the AChE reactor.
3. Results and Discussion
3.1 Inhibition Tests versus Different Amounts of AChE
The optimization of enzymatic biosensor is closely
Current (nA)
Time (Ks)
Fig. 2 Biosensor response after inhibition by CB and OP
compounds in mix: i0-Biosensor response in absences of
pesticides; ia-biosensor response after the interaction
between AChE and the pesticides in mix; ib-biosensor
response after the recovery of the AChE inhibition caused
by OP compounds.
 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
linked to amperometric signal read by electrochemical
transducer. Although a high enzyme activity is required
for reproducible and long-term measurements with
suitable electrode response on acetylcholine, sensitive
inhibitor determinations are favoured at low enzyme
loading. Pesticide detection with an optimized
bi-enzyme sequence electrode involves low esterase
activity but high choline oxidase activity allowing the
signal amplitude to be affected by influencing this
enzyme. Choline biosensor was then tested in order to
evaluate the detection limit, linearity change and
reproducibility. The calibration curve for choline
measurements (Fig. 3) showed linearity (R2 = 0.999)
equal to 0.999 in the range between 0.5 and 1 mM with
a slope of 112.47 nA/mM. The reproducibility of the
choline biosensor was good being equal to 2.42% (n =
3 measurements). These results in particular as regards
the range of sensibility of the biosensor were lower
than other findings [26, 27], because the different
amount of enzyme immobilized on the surface of the
electrode in this study was thirty time lower than those
used in the cited paper. The performances of AChE
biosensors depend on their sensitivity to recognize
small amount of pesticides ranging from 5-20 μg·kg-1
according to Codex Alimentarius MRLs established by
European legislation (Reg. EC No. 149/2008) or
quality control ones used in private industries.
The results presented in Table 1 showed the
R2=0.999
0.5
Choline (mM)
(mM)
Fig. 3 Calibration curve of choline oxidase immobilized on
screen printed electrode.
73
detection limits for Paraoxon and Methomyl using
different unit of AChE. The amount of both pesticides
is able to cause a perceptible inhibition of esterase
activity decrease if a lower enzyme loading is
immobilized on CPG glasses.
The calibration curve of Paraoxon on the base of the
AChE inhibition percentage obtained with 2U of AChE
immobilized on the CPG microreactor was reported in
Fig. 4. The linear range (figure insert) is linear between
5-40 µg·L-1 corresponding to 1.28 × 10-8-1.45 × 10-7
mol·L-1 with a degree of inhibition in the range
6%-35%. By use of bi-enzymatic AChE/Tyr SPE
sensor with Phenylacetate as substrate Andreescu et al.
[8] obtained a detection limit of 5 ×10-3 mg·L-1 (1.82 ×
10-8 mol·L-1). By use of mono enzymatic SPE sensor
(AChE immobilized by cross-linking with GA on
working electrode together with a mediator) Li et al.
[28] obtained a detection limit of 4.95 × 10-5 mg·L-1
(1.80 × 10-7 mol·L-1).
3.2 Calibration Curves of Pesticides Using a 3U AChE
Microreactor
The stability of the bi-enzymatic biosensor was
Table 1 Detection limits for Paraoxon and Methomyl using
different amount of AChE immobilized on CPC glasses.
AChE U CB Metomyl (μg·L-1) OP Paraoxon (μg·L-1)
12 300 300
6 50 50
3 5 20
2 - 5
1.5
(ug/L)
(ug/L)
Fig. 4 Calibration curve of Paraoxon with 2U of AChE.
74 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
tested by measuring the analytical response of 9
acetylcholine chloride (AChCl) injections at 0.5 mM.
The amperometric average signal was 85.04 nA ± 0.87
(n = 9) with a R.S.D relative standard deviation =
1.03%. The calibration plots for Methomyl-Paraoxon
in mix (1:1) and Carbaryl-Fenitrothion in mix, obtained
on the base of different inhibition mechanism of CB
and OP compounds versus AChE activity, were shown
in Figs. 5 and 6 respectively. In the same figures, the
calibration curves of single pesticides were displayed
too. No inhibitions were detected at Paraoxon amounts
lower than 20 μg·L-1 for solutions with single
concentrations of OP. The data obtained for the
different calibration curves highlighted that Paraoxon
inhibition power increased when this was injected in
mix with Methomyl compound: in fact, while no
inhibitions were detected for single Paraoxon amounts
lower than 20 μg·L-1, inhibitions on AChE activity
were recorded from 5 μg·L-1 when two pesticide classes
Fig. 5 Calibration curves of Methomyl and Paraoxon
alone and in mix with 3U of AChE.
Fig. 6 Calibration curves of Carbaryl and Fenitrothion
alone and in mix with 3U of AChE.
were injected in FIA system. A different response was
observed for Methomyl compounds, that showed a
higher inhibition power when injected as single
pesticide versus AChE activity. As regards the
sensibility to detect small amounts of Paraoxon,
Carbaryl and Methomyl the AChE/ChOx biosensor
proposed in this study showed results according to
detection limits detected by Acetylcholinesterases
sensor reported in previous papers [8, 24, 29]. No
studies were found in literature for inhibition power of
Fenitrothion versus AChE electrochemical sensor.
Power inhibition versus AChE activity changes
when different molecules belonging to OP and CB
pesticides get in contact with AChE. The calibration
curves obtained with Carbaryl and Fenitrothion (Fig. 6)
showed higher inhibition percentages than Methomyl
and Paraoxon ones. Moreover, in this case of
Fenitrothion power inhibition versus AChE increased
when injected in mix, unlike Carbaryl, that seems to be
suffering from the simultaneous presence of
Fenitrotion compound. The calibration curves
obtained for the tested pesticides highlighted a
different toxic power versus AChE in function of the
specific pesticide compound and found that a probable
effect exists between two or more pesticide
compounds that increases or reduces the toxic power
observed when the single pesticide compound getting
in contact with AChE. The toxicity of four different
binary pesticides (OP and CB) mixture versus AChE
was recently investigated in vivo by juvenile salmon
survival [30]. The results obtained in this research
showed that the binary pesticides mixture caused a
synergistic AChE inhibition that are in contrast with
those detected in our study. The latter may be probably
explained by the different carrying out of the AChE
inhibition tests, hence from different absorption rate of
the different pesticides in blood.
Our method based on sequential discrimination of
OP and CB pesticides, was able to show that a mix of
these inhibitors, with different inhibition power and
mechanisms (reversible or irreversible), affects the
 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
final enzyme activity which is not due to the linear
combination of inhibitory effects of the single pesticide
alone. Further investigation is needed in order to
understand these mechanisms and effects, which are
strictly related to the toxicological effects of pesticide
content in real samples on insects or humans.
3.3 Biosensor Evaluation with Extra Virgin Olive Oil
Spiked Samples
The bi-enzimatic (AchE-ChOx) amperometric
biosensor was used to analyze spiked (Methomyl and
Paraoxon, 10 and 20 μg·kg-1) EV sample. The effect of
the EV matrix on the AChE activity was calculated
measuring the enzyme activity as affected by the
matrix with respect to that in pure hexane. The matrix
effect was calculated according to Del Carlo et al. [17]
by means of the following formula:
Em% = 100  Pm /Pe (6)
where Pm and Pe represent the inhibition percentage
after the injection of free residue EV diluted with
hexane (1:1) and after the injection of pure hexane
respectively.
Influence of organic matrix on the final activity of
AChE was not negligible since the mean matrix effect
10% (Em% = 110) was higher than the standard
deviation (± 3) of measurement. Since the EV matrix
showed an inhibition effect versus AChE equal to 10%
(Em % = 110) of pure hexan, the matrix effect must be
subtracted from the value registered during the analysis
on real matrices. The data on Methomyl and Paraoxon
recovery present in spiked EV samples obtained using
the % I value in the calibration equation of the
corresponding pesticide in mix showed a good
correlation between expected and obtained data
(Table 2). These results highlighted that the
bi-enzymatic biosensor proposed has a limit of
detection of 5 μg·L-1 (calculated as reported in Ref.
[31]) and is therefore able to detect concentration
values lower than those proposed by the European
legislation, thus representing a good and economical
screening method in food security monitoring.
75
Table 2 Biosensor data registered with spiked EV samples.
Expected value (μg·kg-1) Biosensor result
Methomyl
10 10.8
20 21.7
Paraoxon
10 10.4
20 20.6
4. Conclusions
A screen printed ChOx biosensor coupled with an
AChE CPG microreactor was assembled in a FIA
manifold as extensively described in literature to obtain
a simple and cheap device devoted to quantitative
screening of pesticide multiresidues in EV. Results
suggest that the use of AChE biosensor can be suitable
as an early warning system in screening analysis for
food security monitoring.
The obtained data on Methomyl and Paraoxon
recovery present in spiked EV samples showed to be in
good correlation with expected data. The results
highlighted that the bi-enzymatic biosensor proposed
was able to detect lower concentration than the
accepted maximum residue levels proposed by the
European legislation, thus representing an applicable
screening method in food security monitoring.
Preliminary results showed a probable effect of
competition between different classes of pesticides (OP
and CB) and their different mechanisms of inhibition
for the active site of AChE. For this reason further
studies are in progress on this competition effect and on
validation and quality control procedures stated from
European regulations for pesticide residues analysis in
food and feed.
References
[1] European Commission, Regulation (EC) no 396/2005 of
the European Parliament and of the Council of 23
February 2005 on Maximum Residue Levels of
Pesticides in or on Food and Feed of Plant and Animal
Origin and Amending, Council Directive 91/414/EEC,
Eur-Lex Web Site, http://eur-lex. europa.eu/en/index.htm.
[2] European Commission, Regulation (EC) No 149/2008 of
29 January 2008 Amending Regulation (EC) No.
76 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
396/2005 of the European Parliament and of the Council
by Establishing Annexes II, III and IV Setting Maximum
Residue Levels for Products Covered by Annex I, Eur-Lex
Web Site, http://eur-lex.europa.eu/en/index.htm.
[3] European Commission, Regulation (EC) No. 178/2006 of
1 February 2006 Amending Regulation (EC) N_ 396/2005
of the European Parliament and the Council to Establish
Annex I Listing the Food and Feed Products to which
Maximum Levels for Pesticide Residue Apply, Eur-Lex
Web Site, http://eur-lex.europa.eu/en/index.htm.
[4] M. Dominici, R. Guadagnini, D. Sciarra, E. Ucci, Dossier:
Pesticidi nel piatto [Online], 2008,
http://www.legambiente.eu/scienza/cdoc/schedaDoc.php?
id=4861.
[5] C. Cremisini, S. Di Sario, J. Mela, R. Pilloton, G. Palleschi,
Evaluation of the use of free and immobilised
acetylcholinesterase for paraoxon detection with an
amperometric choline oxidase based biosensor, Anal.
Chim. Acta 311 (1995) 273-280.
[6] S. Hernandez, I. Palchetti, M. Mascini, Determination of
anticholinesterase activity for pesticide monitoring using a
thiocoline sensor, Int. J. Environ. Anal. Chem. 54 (2000)
213-222.
[7] M. Albareda-Sirvent, A. Merkoçi, S. Salvador Alegret,
Pesticide determination in tap water and juice samples using
disposable amperometric biosensors made using thick-film
technology, Anal. Chim. Acta 442 (2001) 35-44.
[8] S. Andreescu, A. Avramescu, C. Bala, V. Magearuy, J.L.
Marty, Detection of organophosphourus insecticides with
immobilized acetylcholinesterase comparative study of
two enzyme sensors, Anal. Bional. Chem. 374 (2002)
39-45.
[9] S. Andrescu, L. Barthelmebs, J.L. Marty, Immobilization
of acetylcholinesterase on screen-printed electrodes:
Comparative study between three immobilization methods
and applications to the detection of OP insecticides, Anal.
Chim. Acta 464 (2002) 171-180.
[10] W. Vastarella, V. Rosa, C. Cremisini, L. Della Seta, M.R.
Montereali, R. Pilloton, A preliminary study on
electrochemical biosensors for the determination of total
cholinesterase inhibitors in strawberries, Int. J. Environ.
Anal. Chem. 87 (2007) 689-699.
[11] S. Viswanathan, H. Radecka, J. Radecki, Electrochemical
biosensor for pesticides based on acetylcholinesterase
immobilized on polyaniline deposited vertically
assembled carbon nanotubes wrapped with ssDNA,
Biosens. Bioelectron. 24 (2009) 2772-2777.
[12] M. Stoytcheva, R. Zlatev, Z. Velkova, B. Valdez, M.
Ovalle, L. Petkov, Hybrid electrochemical biosensor for
organophosphorus pesticides quantification, Electrochim.
Acta 54 (2009) 1721-1727.
[13] A. Guerrieri, L. Monaci, M. Quinto, F. Palmisano, A
disposable amperometric biosensor for rapid screening of
anticholinesterase activity in soil extracts, Analyst 127
(2002) 5-7.
[14] T. Bachmann, R. Schmid, A disposable multielectrode
biosensor for rapid simultaneous detection of the
insecticides paraoxon and carbofuran at high resolution,
Anal. Chim. Acta 401 (1999) 95-103.
[15] M. Cortina, M. Del Valle, J.L. Marty, Electronic tongue
using an enzyme inhibition biosensor array forthe
resolution of pesticide mixtures, Electroanal 20 (2008)
54-60.
[16] B. Gilbert-López, J.F. García-Reyes, A. Molina-Díaz,
Sample treatment and determination of pesticide residues
in fatty vegetable matrices: A review, Talanta 79 (2009)
109-128.
[17] M. Del Carlo, M. Mascini, A. Pepe, G. Diletti, D.
Compagnoni, Screening of food samples for carbamate
and organophosphate pesticides using an electrochemical
bioassay, Food Chem. 84 (2004) 651-656.
[18] N. Mionetto, L.L.Marty, I. Karube, Acetylcholinesterase
in organic solvents for the detection of pesticides:
Biosensor application Biosens, Bioelectron. 9 (1994)
463-470.
[19] I. Palchetti, A. Cagnini, M. Del Carlo, C. Coppi, M.
Mascini, Turner APF determination of anticholinesterase
pesticides in real samples using a disposable biosensor,
Analytica Chimica Acta 337 (1997) 315-321.
[20] L. Campanella, S. De Luca, M.P. Sammartino, M.
Tomassetti, A new organic phase enzyme electrode for the
analysis of organophosphorus pesticides and carbamates,
Anal. Chim. Acta 385 (1999) 59-71.
[21] T.R. Fukuto, Mechanism of action of organophosphorus
and carbamate insecticides, Environ. Health Perspect 87
(1990) 245-254.
[22] A. Guerrieri, G.E. De Benedetto, F. Palmisano, P.G.
Zambonin, Amperometric sensor for choline and
acetylcholine based on a platinum electrode modified by a
co-crosslinked bienzymic system, Analyst 120 (1995)
2731-2736.
[23] I. Palchetti, S. Laschi, M. Mascini, Electrochemical
biosensor technology: Application to pesticide detection,
Meth. Mol. Biol. 504 (2009) 115.
[24] F.N. Kok, V. Hasirci, Determination of binary pesticide
mixtures by an acetylcholinesterase-choline oxidase
biosensor, Biosens. Bioelectron 19 (2004) 661-665.
[25] A.A. Ciucu, C. Negulescu, R.P. Baldwin, Detection of
pesticides using an amperometric biosensor based on
ferophthalocyanine chemically modified carbon paste
electrode and immobilized bienzymatic system, Biosens
Bioelectron 18 (2003) 303-310.
[26] F. Ricci, A. Amine, G. Palleschi, D. Moscone, Prussian
Blue based screen printed biosensors with improved
 Quantitative Screening and Resolution of Carbamic and Organoposphate Pesticides Mixture 77
in Extra Virgin Olive Oil by Acetylcholinesterase-Choline Oxidase Sensor
characteristics of long-term lifetime and pH stability, [29] G.S. Nunes, P. Skladal, H. Yamanaka, D. Barcelo,
Biosens Bioelectron 18 (2003) 165-174. Determination of carbamate residues in crop samples by
[27] I.B. Rejeba, F. Arduinib, A. Arvinted, A. Aminee, M. cholinesterase-based biosensors and chromatographic
Gargouria, L. Micheli, et al., Development of a techniques, Anal. Chim. Acta 362 (1998) 59-68.
bio-electrochemical assay for AFB1 detection in olive oil [30] C.A. Laetz, D.H. Baldwin, T.K. Collier, V. Hebert, J.D.
Biosens, Bioelectron 24 (2009) 1962-1968. Stark, N.L. Scholz, The synergistic toxicity of pesticide
[28] Y.G. Li, Y.X. Zhoua, J.L. Fenga, Z.H. Jiangb, L.R. Mab, mixtures: Implications for risk assessment and the
Immobilization of enzyme on screen-printed electrode by conservation of endangered Pacific salmon, Environl.
exposure to glutaraldehyde vapour for the construction of Health Perspect 117 (2009) 348-353.
amperometric acetylcholinesterase electrodes, Anal. Chim. [31] P.C. Meier, R.E. Zund, SMAC-Statistical Methods in
Acta 382 (1999) 277-282. Analytical Chemistry, Wiley, New York, 1993, p. 103.
Journal of Environmental Science and Engineering A 1 (2012) 78-88
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Nutrient Distribution in Selected Coastal Areas of

Aegean Sea (East Mediterranean Sea)

Alexandra Pavlidou
Hellenic Centre for Marine Research, Anavyssos Attiki 19013, Greece

Received: April 15, 2011 / Accepted: June 14, 2011 / Published: January 20, 2012.

Abstract: The assessment of nutrient levels in various coastal areas of Greece influenced by anthropogenic activities, mainly by
sewage and riverine outflows, is presented in this paper, in order to study the environmental status of the coastal zone of Greece. For
this purpose, the classification scale based on nutrient concentrations, applied especially for the Greek coastal zone, is used. Two
main groups of areas: (a) influenced by agriculture and (b) influenced by domestic and industrial sewage or both agriculture and
sewage have been discerned and treated separately. Two reference sites which are not influenced by anthropogenic activities were
also used. Relatively high nutrient concentrations were recorded near the mouth of the rivers, showing a decreasing trend with
distance. Differences in the proportions of dissolved Si, N, and P in the coastal areas were recorded, which affect the phytoplankton
communities and in some cases have led to the appearance of Harmful Algae Blooms (HABs). Riverine waters determine the level of
nutrient concentrations in the estuaries. The study coastal areas receiving nutrient loads from sewages are, in general, at the same
trophic status and characterized as eutrophic for nitrate and higher mesotrophic for phosphate and ammonium, with phosphate
concentrations between 0.50 and 0.70 μΜ, nitrate ~1.00 μM and ammonium between 1.10 and 2.00 μΜ.

Key words: Nutrient, ratios, environmental status, coastal, Aegean Sea.

1. Introduction continuous environmental pressure derived from a

number of driving forces such as urbanization,
Mediterranean Sea is the largest semi-enclosed
industrialization, changes in land use, tourism
basin in Europe which is separated in smaller basins
development, climate changes etc.. Within only a few
(Alboran Sea, Ligurian Sea, Tyrrhenian Sea, Adriatic
decades, numerous oligotrophic estuarine and coastal
Sea, Ionian Sea, Aegean Sea, Levantine Sea). The
waters have undergone a remarkably troubling
northern margin receives freshwater from a few big
transformation to more mesotrophic and eutrophic
rivers (Ebro, Rhône, Po) and also from smaller ones
conditions, manifested as an increase in phytoplankton
(Axios, Strymon, Nestos, and Evros). Besides the
and macroalgal biomass, increased incidences of toxic
oligotrophic character of the open Mediterranean Sea,
and noxious algal blooms, consequent hypoxia and
elevated nutrient data indicating eutrophication
anoxia, and fish and benthos kills. The variation of
problems occur in some coastal areas of the
nutrient loadings leading to changes in N:P ratio
Mediterranean (e.g. eastern coasts of Spain, gulf of
which controls the planktonic production plays an
Lions, northern Adriatic Sea, Apulian coasts,
important role in the biochemical dynamics and the
Saronikos gulf, Thessaloniki Bay, northern coasts of
functioning of the marine ecosystems.
Greece, etc.) [1]. Coastal areas undergo intense and
Greece is characterized by a high coastal

Corresponding author: Alexandra Pavlidou, associated concentration of population. Indeed, one third of
researcher, Ph.D., main research fields: influence of natural and population is living on a narrow land zone along the
anthropogenic pressures in coastal zone with emphasis in
entrophication and pollution, structure and functioning of Hellenic coastal area. The anthropogenic activities
coastal ecosystems (biogeochemical cycles). E-mail: which influence the natural environment of the
aleka@ath.hcmr.gr.
 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
Hellenic coastal zone are urbanization, industry,
agriculture, aquaculture, tourism, recreational areas
and waste disposal. It is noteworthy that most of the
big cities of Greece are located on the coast and the
urbanized coastal area is estimated around 1.3% of the
total land. Athens and Thessaloniki, the two biggest
cities of Greece where about half of the total
population lives, are located on the coastal zone and
their sewages influence Saronikos and Thermaikos
Gulfs, respectively. Additionally, more than 80% of
the industrial activities in Greece are also located on
the coast and pollute the coastal marine ecosystems by
direct discharges of industrial and domestic wastes.
On the other hand, coastal agriculture covers 35% of
the coastal land enriching the coastal marine
environments in nutrients. Although aquaculture
activities experienced a rapid development during the
last two decades, in general they do not influence
significantly the marine environments, as the fish
farmers choose isolated areas near rocky coasts with
good water mass circulation. Additionally, 90% of
tourism activities are located on the coast and
influence the marine environment [2]. According to
the aforementioned, nutrient input in the coastal zone
of Greece originates mainly from agriculture, riverine
input, municipal wastewaters, aquaculture and tourism
activities. The assessment of nutrient levels in various
Greek coastal areas influenced by anthropogenic
activities, sewage outfalls, riverine outflows etc. gives
the picture of the environmental status of the Greek
coastal zone.
The aims of this paper are to briefly describe the
present status of nutrient levels in selected coastal
areas of Greece. For this purpose, two categories of
areas have been discerned and treated separately
together with reference sites which are not affected by
agricultural or anthropogenic activities. The data used
were obtained during the period 1995-2009.
2. Experiment
For data collection, small boats and the research
79
vessels “Aegaeo” and “Filia” of the Hellenic Center
for Marine Research (HCMR) were used. Seawater
samples were collected from standard depths (surface,
10, 20, 50, 75, 100 m) with Niskin bottles, either
mounted on a rosette or individually on a hydro wire.
The mean integrated values of the water column of
each station were used for all the study areas (only the
layer 0-100 m was used for the calculation of the
mean integrated values). Data concerning Elefsis Bay
(1 station was used), Psittalia sewage outfall (1
station), Inner Saronikos gulf (5 stations),
Thessaloniki Bay (5 stations) and Thermaikos gulf (10
stations) were obtained from four to twelve samplings
per year during 1995 to 2007. For the Outer Saronikos
gulf (Reference site) seasonal data for the period
1995-2004 were used from one station. Data from
2008 and 2009 were used for north Lakonikos (29
stations) and Malliakos gulfs (10 stations), respectively,
whereas, data from 1997-1999 and 2004-2005 were
used for the assessment of nutrient levels in
Strymonikos (20 stations) and Korinthiakos (6 stations)
gulfs, respectively. Finally, for the reference site in the
Aegean Sea, data from 2008 were used.
Seawater samples for nutrients analysis were
collected in 100 mL polyethylene bottles “aged” with
HCl 10% and kept deep-frozen (-20 C) until their
analysis for nitrate, nitrite and silicate, in the
laboratory. The filtered samples taken between 1995
and 1999 were analyzed with an ALPKEM
autoanalyser, from 1999 to 2005 with a
BRAN+LUEBBE II autoanalyser and from
2005-today with BRAN+LUEBBE III autoanalyser
according to standard methods [3, 4]. The
determination of ammonium and phosphate was
performed with Perkin Elmer 20 Lambda
spectrophotometer according to standard methods [5,
6]. Quality assurance of measurements has been
achieved using quality control charts analyzing
certified reference materials (VKI) and participation to
the QUASIMEME inter-laboratory exercise.
Additionally, the lab is certified according to the
80 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
ELOT EN ISO/IEC 17025 procedure.
As the Greek coastal regions differ in the type of
influence, two main groups of areas (influenced by
agriculture, influenced by domestic and industrial
sewage or both by agriculture and sewage) and two
reference sites have been discerned and treated
separately. Strymonikos gulf, Lakonikos gulf, and
Malliakos gulf are some representative coastal areas
mainly influenced by fresh-water (river plumes).
Elefsis Bay, Thessaloniki Bay, Psittalia Sewage
outfall, Korinthiakos gulf and Inner Saronikos gulf are
mainly influenced by domestic and industrial effluents,
whereas Thermaikos Gulf is influenced by both
agriculture and sewage. The outer Saronikos gulf and
the Aegean Sea are considered as unpolluted areas
(Fig. 1).
3. Results and Discussion
3.1 Coastal Areas Influenced by Rivers
It is known that a common characteristic of the
relative small rivers of Greece is their variable
Fig. 1 Map of the studied Greek coastal areas.
behavior between winter/spring and summer/autumn.
High freshwater discharge periods begin generally in
December, following the commencement of the rainy
season, and culminate around May, when snow melts
on the mountains [7]. This seasonality results in
relatively higher flow during the “wet” period
(December-May) and very low flow during the “dry”
period (June-November) affecting the nutrient
distributions. The data used for Thermaikos gulf
contain both wet and dry periods, whereas data for
Strymonikos and Malliakos gulfs concern wet periods
and data for Lakonikos gulf represent the end of the
wet period (data during June). Table 1 presents the
mean integrated concentrations (av) for the water
column (0-100 m) of the four selected areas and
standard deviation (S) for the number (n) of data used.
A decreasing trend of nutrient concentrations from
the river plume is generally observed in all the
selected coastal areas. The nutrient distributions
follow the salinity contours and show that in most
of the cases the influence of the riverine waters is
 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea) 81

Table 1 Mean integrated concentrations and atomic ratios of dissolved inorganic nutrients in Greek coastal areas
influenced by rivers.
Nurtient conc. (μM) Strymonikos Gulf Malliakos Gulf Lakonikos Gulf Thermaikos Gulf
n 60 31 29 630
Nitrate + nitrite av 2.428 1.000 0.260 1.180
S 3.293 1.049 0.270 0.760
n 60 43 29 630
Ammonium av 0.210 0.550 0.258 0.770
S 0.152 0.452 0.105 0.785
n 60 31 29 630
Silicate av 4.292 7.818 1.274 3.699
S 3.646 6.166 0.477 1.833
n 60 31 29 630
Phosphate av 0.090 0.141 0.100 0.335
S 0.063 0.061 0.021 0.177
n 60 31 29 630
N:P av 27.8 11.7 5.48 6.40
S 22.4 9.13 3.41 4.71
n 60 31 29 630
Si:N av 2.29 5.50 2.97 2.28
S 1.11 2.30 1.65 1.45
n 60 31 29 630
Si:P av 49.1 59.0 12.8 13.6
S 20.3 42.8 3.65 14.4
n = number of data; av = mean integrated value; S = standard deviation.

restricted close to the mouth of the river [8]. On the than an urban and/or industrial one. The river enriches
other hand, even the major rivers of Greece (e.g. the marine system of Strymonikos gulf in nutrient,
Axios, Strymon, Evros, Sperchios) influence the indicating a eutrophic environment near the mouth of
coastal marine environments much less when the river (~40 μM for DIN at the surface layer near the
compared to the major rivers outflowing in the mouth of the river). As mixing with offshore waters
Mediterranean Sea and/or in the Black Sea. progressed, nutrient levels declined dramatically (~7.0
The selected areas in the coastal zone of Greece μM in a distance of about 4-5 miles) (Fig. 2).
affected mainly by riverine discharges are briefly
described bellow.
Strymonikos Gulf receives the waters from
Strymon River, which originates in the Vitosha
Mountain in Bulgaria and has total length of about
360 km. The total annual water discharge is around 2
km3 [9]. Strymon River is significantly enriched in
nutrients and considered to be a rather eutrophic river
with nitrate to be the main source of inorganic
nitrogen (nitrate concentration: ~ 150 μM; ammonium:
~ 5.00 μM; nitrite ~1.5 μM) [10, 11]. Strymon River
receives nutrients from the agriculture activities of the
catchment area. According to Pavlidou and
Georgopoulos [9], Strymon River is a P-limited river Fig. 2 Surface distribution of nitrate in Strymonikos Gulf
and can be characterized as an agricultural river rather during March 1999.
82 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
Malliakos Gulf is influenced by Sperchios River
which has 82 km length and mean annual discharge
0.703 km3 [9], which is significantly increased during
wet period. During March and April 2009, massive
deaths of fish in Malliakos Gulf were recorded
coinciding to the high river discharge in the gulf due
to extreme rain events during the previous days.
Lakonikos Gulf is influenced by Evrotas River with
length of 90 km and mean annual discharge 0.76 km3.
The catchment area is mainly used for agricultural
purposes. The industrial activities in Lakonikos
coastal area are very few, so, in terms of pollution
sources, agricultural activities are the most important.
Thermaikos Gulf is considered to be one of the
most polluted coastal zones in Greece. It is the final
receptor of both municipal and industrial wastewaters
from the city of Thessaloniki, through Thessaloniki
Bay, and of two heavily polluted rivers, Axios and
Aliakmon. On the other hand shellfish farming
activity, especially in the Axios Delta, reaches 85% of
the total Greek production. Eutrophication problems
have been also reported in the past [12, 13]. Axios
River originates in Former Yugoslav Republic of
Macedonia and has a total length of 380 km and mean
annual discharge 3.62 km3. Axios river receives
substantial loads of nutrients mainly from agricultural
activities and extended use of N-fertilizers in the
catchment area, but also from the use of detergents
and other point sources [14]. Aliakmon River has total
length 310 km and mean annual discharge 2.7 km3 [9],
whereas, Loudias River has a mean annual discharge
of 0.50 km3.
Substantial differences exist in the nutrient structure
of the four selected Greek coastal areas influenced by
rivers (Fig. 3). It is obvious from Fig. 3 that Evrotas
River has no significant influence in Lakonikos Gulf.
This is probably because the river discharge was
decreased in June. However, Thermaikos seems to be
enriched in phosphate and ammonium, whereas
Malliakos Gulf is much enriched in silicate. It seems
that the loads from Thessaloniki’s waste water
treatment plant affect the region to a greater extent,
southern of Thessaloniki Bay, where the inner
Thermaikos Gulf is located.
On the other hand Strymonikos Gulf is enriched in
nitrate. N:P ratios calculated for the four different
areas revealed interesting differences. Except from
Strymonikos Gulf, in all the other areas the N:P ratio
was calculated lower than the Redfield theoretical one
of 16:1, indicating nitrogen limitation for
phytoplankton growth. In Strymonikos Gulf DIN
concentration is significant high with nitrate to be the
dominant form of inorganic nitrogen. The limiting
factor for phytoplankton growth is Phorphorus (P-
limited). According to these results, phosphorus
controls the functioning of Strymonikos marine
ecosystem, whereas, in Malliakos Gulf silicate plays a
significant role in the functioning of the ecosystem.
This is supported not only by the significantly high
silicate concentration recorded in Maliakos Gulf, but
also by the Si:N and Si:P ratios (Table 1). It is
noteworthy that in Malliakos Gulf the N:P ratio
showed significant spatial variability, indicating a
very fragile ecosystem. It is known that different
species groups display preferences for specific
nutrient regimes, including nutrient ratio or form [15,
16]. For example, diatoms require silicon for their
growth, whereas certain dinoflagellates apparently
have a higher phosphorus requirement than some
other species groups. In March and April 2009, high
concentrations of Chattonella-like sp. were measured
which was assumed to be responsible for the massive
death of fish in Malliakos Gulf. Moreover, elevated
concentrations of diatoms were recorded (e.g.
Pseudonitzschia sp.) and the Karenia brevis sp, in
relatively low concentrations [17]. It seems that
Sperchios riverine waters influence significantly the
functioning of Maliakos Gulf ecosystem.
Thermaikos Gulf is mainly influenced by riverine
inputs, whereas, its northern part is influenced by both
industrial-domestic effluents from Thessaloniki and
riverine inputs. Our data have revealed eutrophication
 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea) 83

Fig. 3 Mean integrated values of nutrient concentrations and their atomic ratios in different coastal areas influenced by
rivers (DIN stands for nitrate + nitrite + ammonium). 0-100 m depth of the water column has been used. The average values
(left axis, bar) and the minimum and maximum values of each parameter (right axis, line) are presented.

problems in the area. Additionally, low N:P ratios
indicate that nitrogen is the limiting nutrient. The
decrease in N:P ratio in Thermaikos Gulf is mainly
due to phosphorus loading probably from detergents
used in FYRUM or/and in reduction of N-fertilizers
used. The decrease in N:P ratio can be related to
increased abundance of certain harmful dinoflagellate
species [18-20]. Indeed, in Thermaikos Gulf, HABs
were recorded frequently. Dinoflagellates were the
dominant blooming species from 1996 and among them
was the toxic species Dinophysis acuminata, a DSP
causative. The first confirmed bloom of Dinophysis
acuminate was recorded from January to May 2000 and
was repeated every year during spring [21, 22].
According to Justic [23], three criteria have been
applied for P-limitation, N-limitation and Si-limitation
as follows: (1) P-limitation, if Si:P > 22 and N:P > 22;
(2) N-limitation, if N:P < 10 and Si:N > 1; (3) Si
limitation, if Si:P < 10 and Si:N < 10. From the
examined coastal ecosystems influenced by rivers,
Thermaikos and Lakonikos Gulfs were found to be
N-limited, whereas, Strymonikos Gulf was P-limited.
The limitation of Si did not found. However,
Malliakos Gulf seems to be a very complicated and
fragile ecosystem, where nutrient dynamics depends
on the Sperchios discharges (Table 1).
84 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
According to the water quality indices applied for
the Greek coastal areas by Karydis [24], the study
areas Lakonikos Gulf is characterized as an
oligotrophic ecosystem, Thermaikos Gulf is eutrophic
for nitrate, upper mesotrophic for phosphate and
oligotrophic for ammonium. Moreover, Strymonikos
Gulf is eutrophic for nitrate and oligotrophic for
phosphate and ammonium. Finally, Malliakos Gulf is
characterized oligotrophic for nitrate and ammonium
and lower mesotrophic for phosphate. For this
classification we have used only the nutrient
concentrations in the study areas.
3.2 Coastal Areas Influenced Mainly by Industrial and
Domestic Sewage
Saronikos Gulf and Thessaloniki Bay receives
effluents from Athens Metropolitan area (population
over 5 million) and Thessaloniki City (population over
1 million), respectively. Thessaloniki Bay is
considered to be polluted due to several urban and
industrial activities such as partially treated municipal
wastewaters, untreated industrial effluents, harbor
activities etc. [25]. Korinthiakos Gulf is affected by
agriculture activities along the coastal zone and
pollutants from domestic and industrial effluents (e.g.
Aluminium Factory in the north coast) [26]. Elefsis
Bay is considered to be a eutrophic area. Significant
number of industries (oil refineries, shipyards,
chemical plants, food, metal cement industries etc.) is
located along the northern coastline of the Bay. The
eastern reach of Elefsis Bay lies in the city of Athens.
Until 1994, the bay was receiving, through Keratsini
channel, the untreated domestic and industrial sewage
of the Athens Metropolitan area, which was
discharged into the surface water layer of the channel
and enriched the bay in metals, nutrients and organic
matter. The aforementioned anthropogenic inputs lead
to ecological stress in the bay and intermittently
anoxic conditions were developed during summer
period every year. After 1994, the sewage of the
Athens Metropolitan area were primarily treated in
Psitallia Sewage Treatment Plant and discharged in
the inner Saronikos Gulf (see Fig. 1). Additionally, in
the end of 2004 the secondary stage of the Psittalia
Sewage Plant was operated [27].
Thessaloniki Bay and Psittalia Sewage outfall
receive the highest load of phosphate, resulting to
significant low N:P ratios, much lower than the
theoretical one (16:1) (Table 2, Fig. 4). This leads to
the existence of toxic species of phytoplankton in
Thessaloniki Bay, as mentioned above. From Fig. 4, it
is obvious that Elefsis Bay, Psittalia sewage outfall
and Thessaloniki Bay are the most polluted among the
study areas. High silicate concentrations were
observed in Elefsis Bay due to remineralisation
processes of the accumulated organic matter during
the anoxic periods. It must be mentioned that diatoms
are the predominant phytoplankton species in Elefsis
Bay (Pagou and Assimakopoulou, unpublished data).
Except from the outer Saronikos gulf, all the other
areas studied are N-limited.
According to the water quality indices applied for
the Greek coastal areas, Elefsis Bay, Thessaloniki Bay
and the Psittalia Sewage outfall are at the same trophic
status and characterized as eutrophic for nitrate and
higher mesotrophic for phosphate and ammonium.
The Inner Saronikos gulf which receives the sewage
from the Psittalia outfall is characterized as higher
mesotrophic for nitrate and lower mesotrophic for
ammonium and phosphate. The water masses
circulation plays an important role in the nutrient
dynamics of the outer Saronikos Gulf.
On the other hand, Korinthiakos Gulf does not seem
to receive significant amounts of ammonium and
phosphate and is characterized as oligotrophic area
according to ammonium and lower mesotrophic
according to phosphate values, whereas is eutrophic
for nitrate. Finally, the outer Saronikos Gulf is
oligotrophic for ammonium, lower mesotrophic for
phosphate and upper mesotrophic for nitrate.
Figs. 5-7 show the ecological quality of the selected
coastal areas of Greece based only on phosphate,
 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea) 85

Table 2 Mean integrated concentrations and atomic ratios of dissolved inorganic nutrients in Greek coastal areas mainly
influenced by sewages.
Nurtient Conc. Elefsis Bay Psittalia Sewage Inner Saronikos Thessaloniki Bay Korinthiakos Gulf
(μM) Outfall Gulf (0-100m)
n 91 98 495 248 23
Nitrate + Nitrite av 1.882 1.201 1.155 1.413 1.752
S 1.401 1.226 0.444 1.342 0.799
n 91 98 495 248 23
Ammonium av 1.120 1.294 0.623 1.059 0.265
S 1.388 1.506 0.562 1.903 0.224
n 91 98 495 248 23
Silicate av 6.250 1.854 1.909 3.990 2.571
S 4.430 0.725 0.520 3.182 0.578
n 91 98 495 248 23
Phosphate av 0.380 0.411 0.259 0.748 0.156
S 0.427 0.556 0.151 1.084 0.020
n 91 98 495 248 23
N:P av 12.8 7.85 7.64 4.32 12.9
S 9.94 6.49 3.36 3.22 5.09
n 91 98 495 248 23
Si:N av 2.44 1.04 1.11 2.56 1.56
S 1.91 0.66 0.40 1.82 1.01
n 91 98 495 248 23
Si:P av 23.4 6.68 8.02 8.59 16.7
S 21.0 4.87 3.65 4.90 4.08
N = number of data; av = mean integrated value; S = standard deviation.
PO4

SO4
NH4

Fig. 4 Mean values of nutrient concentrations and their atomic ratios in coastal areas influenced by industrial and domestic
sewage (DIN stands for nitrate + nitrite + ammonium). The average values (left axis, bar) and the minimum and maximum
values of each parameter (right axis, line) are presented.
86 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)

Fig. 5 Ecological quality of coastal areas in Greece based on phosphate concentrations.

Fig. 6 Ecological quality of coastal areas in Greece based on nitrate concentrations.

 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
Fig. 7 Ecological quality of coastal areas in Greece based on ammonium concentrations.
nitrate and ammonium concentrations, respectively.
White color represents the oligotrophic areas, the light
gray color the lower mesotrophic, dark gray color the
upper mesotrophic and finally the black color
indicates the eutrophic coastal areas of Greece.
According to Karydis [24], water quality classification
is based on phosphate, nitrate and ammonium
concentrations. A coastal area is characterized as:
oligotrophic, with nutrient concentrations < 0.07 μΜ
for phosphate, < 0.62 μM for nitrate and < 0.55 μM
for ammonium; lower mesotrophic, with
concentrations 0.07-0.14 μM for phosphate, 0.62-0.65
μM for nitrate and 0.55-1.05 μM for ammonium;
upper mesotrophic, with concentrations 0.14-0.68 μM
for phosphate, 0.65-1.09 μM for nitrate and 1.05-2.2
μM for ammonium and eutrophic, with concentrations
higher than 0.68 μM for phosphate, 1.19 for nitrate
and 2.20 for ammonium.
In general, the south part of Greece is less eutrophic
than the north part. Additionally, the coastal area of
Greece is characterised by relatively low ammonium
87
concentrations, whereas, according to nitrate and
phosphate values is characterized more eutrophic.
Phosphate and nitrate seem to control the functioning
of the coastal ecosystems of Greece.
4. Conclusions
Differences in the proportions of dissolved Si, N,
and P in the coastal areas were recorded, which affect
the phytoplankton communities and in some cases
have led to the appearance of HABs. Riverine waters
determine the level of eutrophication in the estuaries.
In general, enrichment caused by riverine nutrient
loads is restricted close to the mouth of the river. On
the contrary, the coastal areas receiving nutrient loads
from sewages are, in general, at the same trophic
status and characterized as eutrophic for nitrate and
higher mesotrophic for phosphate and ammonium.
References
[1] A. Pavlidou, E. Souvermezoglou, Nutrients in the
Mediterranean Sea, in: Pollution in the Mediterranean Sea:
An Overview from IASON SSA, Hellenic Centre for Marine
88 Nutrient Distribution in Selected Coastal Areas of Aegean Sea (East Mediterranean Sea)
Research (HCMR), Athens, Greece, 2005, pp. 5-12.
[2] Ch. Anagnostou, V. Drakopoulou, V. Kapsimalis, E.
Papathanasiou, Anthropogenic activities along the
Hellenic Coasts, in: State of the Hellenic Marine
Environment, HCMR, Athens, Greece, 2005, pp.
280-284.
[3] F. Koroleff, Revised version of “direct determination of
ammonia in natural waters as indophenol blue”, in: Int
Con Explor Sea C M 1969/C:9 ICES Information on
Techniques and Methods for Sea Water Analysis, Interlab
Rep, No. 3, 1970, pp. 19-22.
[4] J.B. Mullin, J.P. Riley, The colorimetric determination of
silicate with special reference to sea and natural waters,
Anal Chim Acta 12 (1955) 162-176.
[5] J.D.H. Strickland, T.R. Parsons, A practical handbook of
sea water analysis, Bull Fish Res Bd Canada 167 (1968)
310.
[6] J. Murphy, J.P. Riley, A modified single solution method
for phosphate in natural waters, Anal Chim Acta 27
(1962) 31-36.
[7] A. Karageorgis, W. Gardner, D. Georgopoulos, A.
Mishonovs, E. Krasakopoulou, Ch. Anagnostou, Particle
dynamics in the Eastern Mediterranean Sea: A synthesis
based on light transmission, PMC, and POC archives
(1991-2001), Deep Sea Research Part I 55, 2008, pp.
177-202.
[8] A. Pavlidou, Dissolved oxygen and nutrient
concentrations in the Inner Thermaikos gulf, in:
Monitoring of Thermaikos Gulf Marine Ecosystem,
Technical report, HCMR, Athens, Greece, 2005.
[9] N. Skoulikidis, Environmental state of rivers in the
Balkans—A review within the DPSIR framework,
Science of the Total Environment 407 (2009) 2501-2516.
[10] L. Ignadiades, M. Karydis, P. Vounatsou, A possible
method for evaluating oligotrophy and eutrophication
based on nutrient concentrations scales, Mar. Pol. Bull.
24 (1992) 238-243.
[11] A. Pavlidou, D. Georgopoulos, Dissolved oxygen and
nutrients in coastal waters impacted by the Strymon River
plume, North Aegean Sea, Greece, Global Nest 3 (2001)
71-84.
[12] O. Gotsis-Skretas, N. Friligos, Contribution to
eutrophication and phytoplankton ecology in the
Thermaikos Gulf, Thalassographica 13 (1990) 1-12.
[13] A. Giannakourou, T.Y. Orlova, G. Assimakopoulou, K.
Pagou, Dinoflagellate cysts in recent marine sediments
from Thermaikos Gulf, Greece: Effects of resuspension
events on vertical cyst distribution, Continental Shelf
Research 25 (2005) 2585-2596.
[14] A.P. Karageorgis, N.P. Nikolaidis, H. Karamanos, N.
Skoulikidis, Water and sediment quality assessment of
Axios River and its coastal environment, Continental
Shelf Research 23 (2003) 1929-1944.
[15] J. Heisler, P.M. Glibert, J.M. Burkholder, D.M. Anderson,
W. Cochlan, W.C. Dennison, et al., Eutrophication and
harmful algal blooms: A scientific consensus, Harmful
Algae, Aug. 2008, pp. 3-13.
[16] A. Pavlidou, Dissolved Study for the Investigation of the
Massive Fish Deaths in Malliakos Gulf in March and
April 2009, HCMR, Athens, Greece, 2009, pp. 9-22. (in
Greek)
[17] K. Pagou, Study for the Investigation of the Massive Fish
Deaths in Malliakos Gulf in March and April 2009,
HCMR, Athens, Greece, 2009, pp. 26-44. (in Greek)
[18] S. Reizopoulou, E. Strogyloudi, A. Giannakourou, J.
Hatzianestis, I. Varkitzi, E. Graneli, Accumulation of
Okadaic acid in shellfish from Thermaikos Gulf (Eastern
Mediterranean), in: Proceedings of 11th International
Conference on Harmful Algae, 2004.
[19] K. Koukaras, G. Nikolaidis, Dinophysis blooms in Greek
coastal waters (Thermaikos Gulf, NW Aegean Sea), J.
Plankton Res. 26 (2004) 445-457.
[20] Κ. Pagou, G. Assimakopoulou, Th. Zoulias, Study of
phytoplankton biomass and communities, in: Monitoring
of the Marine Environment of Thessaloniki Ba, HCMR,
2001, pp. 113-136.
[21] K. Pagou, Eutrophication in Hellenic coastal areas, in:
State of the Hellenic Marine Environment, HCMR,
Athens, Greece, 2005, pp. 127-136.
[22] I. Varkitzi, K. Pagou, I. Hatzianestis, C. Pyrgaki, A.
Pavlidou, G. Asimakopoulou, et al., Effects of nutrients
availability on the growth and toxicity of the
dinoflagellate Prorocentrum lim, in: Proceedings of 8th
Symposium of Oceanography and Fisheries, 2006.
[23] D. Justic, N. Rabalais, R. Turner, Q. Dortch, Changes in
nutrient structure of river-dominated coastal-waters:
stoichiometric nutrient balance and its consequences,
Estuarine, Coastal and Shelf Science 40 (1995) 339-356.
[24] M. Karydis, Evaluation of the Trophic Levels in Greek
Coastal Ecosystems, Scientific report, Univ. of Aegean,
1999.
[25] A. Pavlidou, R. Psyllidou Giouranovits, Trophic
conditions and trends in nutrient concentration in
Thessaloniki Bay, in: Proceedings of the 37th CIESM
Congress, Vol. 37, 2004, p. 232.
[26] A. Pavlidou, DO and Nutrient Concentrations in
Korinthiakos Gulf during 2004-2005, Final technical
report, HCMR, 2006.
[27] A. Pavlidou, H. Kontoyiannis, R. Psyllidou-Giouranovits,
Nutrient concentrations in the Inner Saronikos Gulf after
the sewage secondary treatment, Fresenius
Environmental Bulletin 17 (2008) 1852-1858.
Journal of Environmental Science and Engineering A 1 (2012) 89-98
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Possibilities of Biorefinery by Sweet Sorghum

(Sucrosorgo) on Brownfield Site

Ágnes Dergez1, Diána Bordás1, Helga Kovács2, Gergely Forrai2, György Zsigrai3, Sándor Szabó3, Balázs Magyar4
and István Kiss1
1. Bay Zoltán Foundation for Applied Research Institute for Biotechnology, Szeged H-6726, Hungary
2. Bay Zoltán Foundation for Applied Research Institute of Logistics and Production Engineering, Miskolc H-3519, Hungary
3. Karcag Research Institute, University of Debrecen, Debrecen H-5300, Hungary
4. Elgoscar-2000 Ltd., Budapest H-1134, Hungary

Received: January 26, 2011 / Accepted: March 30, 2011 / Published: January 20, 2012.

Abstract: There are several thousand hectares of so-called brownfield sites in Europe, which are not suitable for agricultural use due to
industrial pollution. Green biorefinery technologies allow economical reutilization of hydrocarbon or heavy metal contaminated sites
by growing agricultural plants and herbs, which are able to tolerate pollutions and produce high amount of biomass or can be sources of
valuable products. The applicability of biorefinery was tested on a heavy metal polluted soil. Complete by-product utilization of the
biomass was aimed for energetic use and for the production of precursors for chemical industry from plants cultivated on brownfield
site. Near hundred species and varieties were cultivated and tested for potential utilizable components. In the present study sweet
sorghum was chosen for testing biorefinery, because of its high sugar content and tolerance against heavy metals. After pressing, the
liquid fraction was tested as potential raw material for bioethanol production in a pilot system, and the cellulose-rich solid fraction was
investigated for biogas and thermal energy production. Our results show that biorefinery is a real possibility for the utilization of
brownfield sites for industrial purposes. Furthermore, we established a technology that ensures the biomass cycle and the utilization of
solar energy bound in plants.

Key words: Biorefinery, brownfield, heavy metal pollution, phytoremediation.

1. Introduction These underused and contaminated sites have been

threatening the environment including the surface and
The definition of brownfield sites is not consistent in
subsurface water resources. Additionally,
Europe; however, the CABERNET (Concerted Action
environmental pollution generally hampers the
on Brownfields and Economic Regneration) generally
utilization of brownfields. In most cases, these
defines them as “sites that have been affected by the
pollutions formed anthropogenically by industrial
former uses of the site and surrounding land; are
activity. The contaminants have very diverse chemical
derelict and underused; may have real or perceived
composition: like paraffins and olefins, cyclic paraffins
contamination problems; are mainly in developed
and olefins, polyaromatic hydrocarbons originating
urban areas; and require intervention to bring them
from crude oil, or heavy metals from mining and
back to beneficial use”. Economical utilization of
halogenated compounds, solvents, xenobiotics,
brownfield sites is a great problem all over the world;
pesticides originating from chemical industry and
however, there are several thousand hectares of
agriculture [2]. Heavy metals mostly originate from
brownfields in Europe [1].
mining, metallurgy and traffic. The heavy metal
pollution changes the microbiological composition of
Corresponding author: Ágnes Dergez, M.Sc., research the soil, which has a detrimental effect on soil quality
fields: bioremediation, biorefinery. E-mail: dergi@baybio.hu.
90 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site
[3]. Heavy metal pollution is a serious environmental
threat in Europe, however, the expansion and the
degree of the contamination is variable. For example,
in Western Europe approximately 1,400,000 sites were
affected by heavy metals, of which, over 300,000 were
contaminated, i.e., the total heavy metal content is
higher than the environmental limit value [4].
Biorefinery is a complex technology, with the aim to
utilize biomass by conversion to useful materials (e.g.
fuels, solvents, plastics, cosmetic and pharmaceutical
materials and food for humans) and/or energy carriers.
The produced energy can cover the energy
consumption of the technology partially or totally.
By-products and the waste streams reducing can
contribute to the sustainable development [5, 6].
During biorefinery, the first step is the separation of
different plant organs and the physically and chemical
fractioning of them by cracking, milling, pressing and
extracting to produce primary products (carbohydrates,
vegetable oils, volatile oils, enzymes, antioxidant
compounds, pigments etc.). Primary products convert to
secondary products chemically or biotechnologically.
Secondary products can be for example solvents,
lubricants, surfactants, bio-plastics, dyes, cosmetics,
pharmaceutical agents, etc.. Finally, residues utilized by
energetically producing biogas or pellets and briquettes
can form electricity or heat energy [7, 8].
Use of agricultural lands to produce non-food plants
(e.g. for the energy or chemical industry) causes social
resistance currently. In contrast, the case is different
with the above-mentioned brownfields, where classical
agricultural output is not feasible due to the quality of
soil and the high level of contaminants, such as heavy
metals. In this case, plant biorefinery may be a possible
method of utilization, since heavy metal tolerant plants
have been described [9, 10]. Heavy metal uptake
differs between and within plant species. Accumulators
concentrate heavy metals in their aerial parts;
indicators regulate the uptake and transport of metals to
the shoot, thus internal concentration reflects the
external levels, at least until toxicity occurs. Excluders
can grow on heavy metal contaminated sites, however
they maintain the concentration in the shoot at low
levels [11].
Consequently, growing plants for biorefinery is a
possibility to utilize brownfields beneficially. However,
plants, which can tolerate, but do not accumulate the
contaminants, have to be chosen. Since the area of the
brownfield sites is usually limited, the applied plants
have to yield high biomass (e.g. energy crops) or
contain a high value material (e.g. medicine herbals),
which enable an economically sustainable cultivation
and biorefinery.
Sorghum and sunflower can take up heavy metals to
their shoot. However, accumulation is limited in plants
by phytotoxicity, therefore, metal content is still at safe
levels for humans and animals [12]. Moreover, plants
that contain high amounts of carbohydrates, such as
sorghum or millet, ethanol or lactic acid can be
produced [13]. Bioethanol is a valuable industrial raw
material, biofuel, solvent [6]. Lactic acid is used in the
food, textile, and chemical industries and it is a raw
material to produce polylactate, which is a
biodegradable plastic [14].
The aim of present paper is to show that biorefinery
is a proper technology to profitable utilization of
brownfields and sweet sorghum, which is a good plant
to produce high-level biomass on heavy metal
contaminated site. Furthermore, we presented a
pilot-scale system to produce bioethanol from the
liquid phase and we investigated the energetically
utilization of the solid phase of sorghum.
2. Materials and Methods
2.1 Description of the Experimental Site
A refuse heap with extremely high-level of heavy
metal content was gathered at the Mátra Metal Mines
(Gyöngyösoroszi, Hungary) due to local ore
enrichment processes. Biomass production capacity,
heavy metal uptake and potential applicability for
chemical industry and energetic utilization of sweet
sorghum (Sucrosorgo) were investigated on a trial plot
 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site
of 3,000 m2, which was established on a moderately
heavy metal polluted soil near the Toka stream, in its
catchment area. TERRASOL compost of 30 t·ha-1 dose
and MAP (mono-ammonium phosphate) of 200 kg·ha-1
dose were applied to satisfy the nutrition demand of
plants.
At the beginning of blooming, plant samples were
taken from 1 m2 areas of the plant standings to establish
the biomass production and to investigate the
possibility of biorefinery of the biomass.
2.2 Heavy Metal and Soil Parameter Analysis
The plant and soil samples were prepared by aqua
regia and Lakanen-Erviö method and the heavy metal
content was determined by ICP-OES according to
MSZ 21470-50:2006 and MSZ EN 13657:2003
standards.
Other soil parameters, such as pH, ammonium,
nitrate, phosphate, were determined by MSZ
21470-2:1981.
2.3 Arrangement of the Pilot System
Based on our previous results and conceptions [15, 16],
Elgoscar-2000 Ltd. has developed a pilot-scale
biorefinery manufactory in Gyöngyösoroszi. The
design of the arrangements was visualised by CAD
program (inventor) [17]. A controller and data
collecting system is connected to the manufactory with
the following parameters: RS232/485 data collecting
net, 9600 baud 6 digital unit, 20 A/D channels (10 bit),
2 impulses-register, saving data per 10 s, 12 MB per
day. The test factoring was carried out in three
independent series.
2.4 Monitoring of Fermentation
Shoots of the plants were physically crushed and
fractioned by a fruit press. 200 L of the liquid phases
were incubated in a fermentor at 20 °C for 4 days after
inoculation of Saccharomyces cerevisiae to convert the
sugar content of plant extracts to ethanol. The liquid
phases were not sterilized before the fermentation and
91
the inoculum was grown in DSM-186 medium and
incubated at 20 °C for 72 hours. 2,000 mL inoculum
was applied in the 200 L volume.
During the fermentation 1 mL liquid phase sample
was taken from the fermentor and the daily
carbohydrates, organic acids and ethanol content were
measured by HPLC. The samples were centrifuged at
13,000 rpm for 20 min and diluted to 20-fold in 0.01 N
H2SO4 eluent and centrifuged again for 20 min (13,000
rpm). The supernatant was analyzed by GynkoTek
HPLC with isocratic elution, on Transgenomic ICSep
Coregel 87H3 column and at 30 ºC, and the eluent was
0.01 N H2SO4 Carbohydrates (sucrose, glucose,
fructose), organic acids (acetic acid, lactic acid), and
ethanol were detected by refractive index detector and
the organic acids and ethanol were detected by UV
detector, too. Data were integrated by the Chromeleon
software.
2.5 Energetical Utilization of the Biomass
2.5.1 Sample Preparation
After crushing and fractioning, the solid phase of
sweet sorghum (Sucrosorgo) was tested for biogas
production and analyzed of its combustion
characteristics. The plants samples were dried at
105 °C for 2 hours then the dried mass were
determined. Dried solid phase for biogas production
and wet solid phase for combustion characteristics
were investigated.
2.5.2 Biogas Production
In our experiments, the reactor volume was 1,000
mL, which contained 800 mL anaerobe sludge and 20 g
dried solid. The reactors were incubated at 37 °C. The
produced biogas were trapped and measured for a
month. The methane content of the biogas was
investigated continuously. Methane content was
determined by GC-MS (Gaschromatograph with 5975
MS Detector). Parameters of the measurement were:
front inlet: 250 °C; split rate: 1:50; GS-GASPRO (60m
× 0.320mm) column; at 40 °C for 6 min, heating to
130 °C by 30 °C/min; at 130 °C for 2.0 min; method:
92 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site

C1-C2GAS.M.; detecting in AUTO tune mode; carrier Hungary), which is the reservoir of contaminants of
gas: He (Linde). Toka stream, and its catchment area. The refuse
2.5.3 Combustion Experiments contains high amounts of heavy metals. We found that
The goals of the measurements include the cadmium, copper, lead, selenium and zinc
determination of the combustion characteristics of the concentrations are 7-17-fold of the environmental limit
sample. The combustion characteristics include value (Table 1). In the experimental area, the heavy
measurement of moisture, volatiles, ash, chemical metal contens are under the environmental limit value,
composition and calorific values. During our however, this site is continuously polluted by the Toka
examinations, we determined the total moisture content stream, which comes from the closed mine area. Heavy
of the sample with a cascade drying according to MSZ metal contents were investigated by not only the aqua
(Hungarian Standards) 24000-23:1977 standard and regia method, but also Lakanen-Erviö method was
determined the ash content according to the MSZ ISO applied, because it can characterizes better the available
1171:1993 standard. We carried out the determination metal contents for plants and amount of metals, which
of the volatile content according to the regulations of can be leached by the stream. In most cases, the heavy
the MSZ 24000-10:1983 standard. According to the metal contents of the refuse approach or pass the limit
MSZ, ISO, DIN, or CEN standards referring to the values (The limit values are regarded to the aqua regia
instrumental determination of carbon, hydrogen, method). According to our results the bioavailable
nitrogen and sulfur content, we carried out our metal content is not high on the experimental site,
measurements with a component analyzer. Following however this is a brownfield, due to the continuous
the referring standards, we determined the contamination and other soil features, for example
caloricity/combustion heat. We carried out 3-3 parallel slightly acidic pH (6.16  0.35), low amounts of
measurements of the sample. ammonium (10.90  6.94 mg·L-1), nitrate (0.84  0.86
mg·L-1), phosphate (0.72  0.30 mg·L-1) ions and the
3. Results and Discussion
history and localization of the site. The low nitrogen and
3.1 Description of the Experimental Site phosphorous contents of the soil made the fertilization
We measured the heavy metal content of the hoarded necessary for the efficient crop production, therefore we
refuse at the Mátra Metal Mines (Gyöngyösoroszi, applied the TERRASOL compost and MAP.
Table 1 Heavy metal (HM) content of the experimental site and the refuse heap, which is the reservoire for contamination.
Experimental site at the catchment area of Heavy metal content of the refuse heap Limit value
Toka stream near the refuse heap (mg·kg-1)a
Heavy metal Heavy metal
HM Total heavy Total heavy
content (mg·kg-1) content
metal content -1 metal content
(by S.D. S.D. (mg·kg ) (by S.D. S.D.
(mg·kg-1) (by (mg·kg-1) (by
Lakanen-Erviö Lakanen-Erviö
aqua regia method) aqua regia method)
method) method)
Ag 0.17 0.01 <0.01 - 0.28 0.13 1.55 0.35 2
Cd 0.28 0.19 0.48 0.09 9.63 5.31 16.26 8.63 1
Co 2.44 0.65 17.70 0.46 3.08 1.20 19.48 3.77 30
Cu 18.60 3.08 58.53 4.42 498.17 100.70 1260.33 301.21 75
Fe 472.67 320.96 45583.33 376.08 1865.50 828.46 91816.67 14536.85 -
Ni 1.45 0.16 18.87 1.48 0.70 0.13 10.54 1.09 40
Pb 12.07 4.03 59.40 11.74 176.57 79.10 700.83 188.56 100
Se 0.12 0.01 0.78 0.04 0.26 0.05 23.53 3.55 1
Zn 20.13 7.85 179.00 11.79 794.67 300.49 3432.17 1677.19 200
a
Source from Ref. [18].
 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site 93

3.2 Biomass Yield and Heavy Metal Uptake of Sweet Table 2 Heavy metal (HM) contents of sweet sorgum
Sorghum (Sucrosorgo) (Sucrosorgo) grown on the brownfield experimental site
(Gyöngyösoroszi) and on agricultural site (Karcag).
We measured the biomass production capacity and
Plants grown on the Plants grown on
heavy metal uptake of sweet sorghum grown on the bownfield site conventional agricultural
brownfield site. The plants were not growing uniformly, HM Gyöngyösoroszi site Karcag
because of the extremely rainy year of 2010 in Hungary. HM content HM content (mg/kg
S.D. S.D.
The regular rainfall retarded both sowing and plant (mg/kg dry mass) dry mass)
growth. After all, at the beginning of blooming or yield Ag 0.56 0.13 0.49 0.26
Cd 5 0.71 0.35 0.04
ripening plant samples were taken. We established that
Co 0.10 0.04 0.07 0.01
the biomass yield was generally 53 ± 3 t·ha-1. Based on Cu 18.40 0.57 9.07 4.29
the normal productivity of the applied sweet sorghum Ni 1.03 0.13 0.54 0.12
variety (80-100 t·ha-1), its estimated productivity would Pb 1.09 0.13 0.33 0.05
Se 0.05 - 0.03 0.00
be 60-70 t·ha-1 in normal weather conditions. After
Zn 165 8.49 40.90 13.72
pressing, about 40% liquid phase formed depending on
the wet mass. Based on our observations, the biomass Table 3 Heavy metal (HM) content of solid and liquid
phases of sweet sorgum (Sucrosorgo) grown on the
production of this plant can be increased by the
brownfield experimental site after pressing.
improvement of the agro-ecological conditions (water
Plants grown on the bownfield site Gyöngyösoroszi
regime, organic matter content, nutrient supply, etc.) of HM content of HM content of
HM solid phase liquid phase
the brownfield sites. S.D. S.D.
(mg/kg dry (mg/kg dry
The heavy metal contents of plants grown on the mass) mass)
experimental site were compared to plants grown on an Ag 0.14 0.00 0.0032 0.0030
Cd 3.62 0.59 0.1242 0.0351
unpolluted agricultural site (Karcag, Hungary). The
Co 0.41 0.05 0.0020 0.0001
results were referred to the dry mass, which was 24.81
Cu 10.46 4.59 0.3000 0.2645
± 0.04% (23.47 ± 2.45% in case of plants originating Ni 29.50 13.01 0.0573 0.0178
from Karcag) in the total plants and 33.95 ± 0.50% in Pb 1.88 0.01 0.0204 0.0028
the solid phase. The results show that cadmium, lead Se <0.02 - 0.0005 0.0001
Zn 124.50 9.19 4.3720 3.6529
and zinc content of plants grown on the brownfield
experimental site were high (Table 2), in spite of that variety. This result indicates the possibility of
the heavy metal content was not higher in the soil than utilization of this variety for phytoremediational
the environmental limit value. purposes to maintain the low heavy metal
Moreover, we investigated the distribution of heavy concentration in the soil in spite if the regular
metals between solid and liquid phases after pressing contamination by the Toka stream.
(Table 3). Results show that most part of the heavy
3.3 Arrangement of the Pilot System
metals were present in the solid phase, therefore heavy
metals may cause problems during combustion and A pilot-scale biorefinery manufactory was
biogas production. established in Gyöngyösoroszi, Hungary. The system
Heavy metal recovery was calculated based on has involved a Bosch Axt 2000 Rapid shredder, an
biomass-yields and the measured heavy metal contents, modified Aquapress Garda (80 L, 6 bar), a fermentor
similarly to heavy metal distribution. However, we (200 L) with an engine to ensure the mixing (0-100 rpm,
would like to highlight that approximately 1 kg·ha-1 1500 W), a heating and cooling cycle with a Hajdu type
zinc was recovered in the Sucrosorgo sweet sorghum water heater and a spooler (100 L), and an Agromex
94 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site
KZK60 distillatory (60 L, propane-butane gas heater)
with a cooling cycle (100 L + 275 L spooler) (Fig. 1).
The liquid flow in the system is moved by a Rover 20
CE type peristaltic pump. The CAD design of the
planned pilot birorefinery system is shown in the Fig. 2.
A controller and data collecting system monitors the
following parameters: temperature of the liquid and air
phase of the fermentor; the temperature, current
consumption and quantity of the circulated water in the
heating cycle; temperature of the liquid and the air
phase in the distillatory; the temperature, current
consumption and quantity of the circulated water in the
cooling cycle; and the operating time of the pumps. The
temperature of the liquid phase is regulated in the
fermentor. It is hold between 16 and 23 °C.
In this system, we could produce ethanol; however,
acidic fermentation is started spontaneously after
shredding and pressing (Fig. 3). The effectiveness of
the pressing was 41.41% ± 0.07.
The controlling system worked appropriately, the
temperature of the liquid phase in the fermentor
changed between 15-22 °C, and temperature of the
Fig. 1 The scheme of the planned pilot birorefinery system.
Fig. 2 The CAD design of the planned pilot birorefinery
system.
distillation was 78-80 °C. However, a mixing engine
was connected to the fermentor, it was not operated to
maintain of anaerobic conditions for fermentation. The
current consumption of the system was 48.7 ± 8.7 kWh
(i.e. 175,440 ± 31,323 kJ) in a fermentation cycle (4
day). The produced ethanol, the spontaneously
produced lactic and acetic acid and the sugar (sucrose,
glucose, fructose) contents were monitored daily.
3.4 Monitoring of the Fermentation
In the pilot-scale biorefinery system, 200 L liquid
phase of the sorghum was transferred to the fermentor.
The liquid was inoculated by yeast (Saccharomyces
cerevisiae) and incubated for 4 days at optimal
temperature for growth (15.24-21.62 °C). During the
fermentation, carbohydrates, organic acids and ethanol
content were followed by HPLC.
The results show that sucrose and glucose was
decreased during the ethanol production (Fig. 3). The
sugar content of the liquid phase of the sorghum was
low (1.84-3.55 g·L-1 sucrose, 4.04-11.94 g·L-1 glucose),
it was converted to ethanol, in spite of that spontaneous
acidic fermentation was observed during the shredding
and pressing. The spontaneous acidic fermentation
produced 0.36-2.41 g·L-1 acetic acid and 1.95-9.91 g·L-1
lactic acid, which turned the pH to 4.48-5.83. The acetic
and lactic acid content did not change in the 2nd and 3rd
(F2, F3) fermentation phases, however it increased in
the first (F1). The reason of this phenomenon may be
that oxygen was dissolved into the liquid after mixing
during the sampling. Based on these results, mixing was
omitted in the 2nd and 3rd (F2, F3) fermentation. On
this pH and the conditions above 3.36-8.48 g·L-1
ethanol was produced. From this material 12.5 L
purified ethanol was produced by distillation having
20.00 ± 4.08% ethanol. In a second distillation step, 6 L
ethanol was produced with 40% ethanol content.
Consequently, sweet sorghum is an appropriate raw
material for ethanol production. However, the
shredding and pressing must be further optimized to
inhibit the spontaneous acidic fermentation. Probably,
cooling and increasing the pH might be appropriate
 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site 95

Glucose content (g·L-1)

4 15
Sucrose content (g·L-1)

3.5
12
3
2.5 9
2 6
1.5
1 3
0.5 0
0 0 1 2 3 4
0 1 2 3 4
time (day)
time (day)

F1 F2 F3 F1 F2 F3

(a) (b)

15 15
Fructose content (g·L-1)

Ethanol content (g·L-1)

12 12
9 9
6 6
3
3
0
0
0 1 2 3 4
0 1 2 3 4
time (day)
time (day)

F1 F2 F3 F1 F2 F3

(c) (d)

15 4
Lactic acid content (g·L-1)

Acetic acid content (g·L-1)

3.5
12
3
9 2.5
2
6 1.5
1
3
0.5
0 0
0 1 2 3 4 0 1 2 3 4
time (day) time (day)

F1 F2 F3 F1 F2 F3

(e) (f)
Fig. 3 Changes of the (a) sucrose, (b) glucose, (c) fructose, (d) ethanol, (e) lactic acid and (f) acetic acid content during the
three independent (F1, F2, F3) fermentations.
96 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site

techniques to solve this problem. It is important to note the same period.

that the spontaneously produced lactic acid could be According to our results, the solid phase of sweet
also a valuable product of biorefinery. sorghum grown on brownfield site is utilizable for
biogas production, and the energy obtained from this
3.5 Energetic Utilization
biogas may supply the energy consumption of the
The solid phase of sweet sorghum (Sucrosorgo) was biorefinery system. Since special equipments are
tested for biogas production and its combustion necessary for biogas utilization (for example gas
characteristics were analyzed. The results were engine with specific settings), we also investigated the
referred to dry mass in case of the biogas production, parameters of combustions of sweet sorghum.
and wet mass in case of combustion experiments. 3.5.2 Combustion Experiments
3.5.1 Biogas Production In the solid phase of sweet sorghum, significant
20 g dried solid phases were mixed with 700 mL energy is available. Therefore, combustion
anaerobic sludge. The produced biogas and methane characteristics which include measurement of moisture,
content were measured for 30 days. Since the most volatiles, ash, chemical composition and calorific
heavy metal was present in the solid phase, we values (Table 4) were determined to find the critical
compared the results to solid phase of sweet sorghum parameters of this plant (herbaceous, grown on
(Sucrosorgo) grown on agricultural soil (Karcag). The brownfield site) during the combustion.
results show that the added sorghum samples increased The results referred to the wet plants. The results
the biogas production significantly and the heavy metal show that carbon content is lower, while nitrogen and
content did not influence the produced biogas (Fig. 4). ash is higher than the same parameters of ligneous
During the biogas production, the methane content plants. Other parameters, like volatile, sulfur, hydrogen
reached approximately 40% after a week and about and the calorific values were consistent with ligneous
60% after a month. The methane content of the biogas plants. The calorific values (combustion heat and
was equalized (i.e. added sorghum samples does not calorimetric value) are similar to that of ligneous plants;
significant effect on methane production), however, therefore, this herbaceous plant was as efficiently
having regard to the higher amount of biogas we could utilizable for heat (energy) production as ligneous
produce approximately 4-fold amount of methane in plants [19]. Moreover, 11.7  2.1 kg solid phase of

4500 70
total amount of biogas (mL)

4000 60
methane content (%)

3500 50
3000
40
2500
30
2000
1500 20

1000 10
500 0
0 7 9 13 20 23 30
0 100 200 300 400 500 600 700 time (day)
time (h)
anaerobic sludge
sweet sorghum (Sucrosorgo) grown on the brownfield site (Gyöngyösoroszi)
sweet sorghum (Sucrosorgo) grown on the brownfield site (Gyöngyösoroszi)
basic activity of the anaerobic sludge
sweet sorghum (Sucrosorgo) grown on the agricultural site (Karcag) sweet sorghum (Sucrosorgo) grown on the agricultural site (Karcag)

(a) (b)
Fig. 4 Biogas production of sweet sorghum (Sucrosorgo) grown on the brownfield site in a monthly experiment compared to
the anaerobic sludge and sweet sorghum (Sucrosorgo) grown on an agricultural site: (a) total amount of biogas; (b) methane
content of the biogas.
 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site
Table 4 Combustion characteristics of sweet sorghum grown on brownfield site.
Moisture Ash Volatile C H
(m/m%) (m/m%) (m/m%) (m/m%) (m/m%)
19.3 4.7 63 36.34 4.96
sorghum is enough to cover the energy consumption of
the pilot-scale biorefinery system during an ethanol
fermentation cycle and it can be cultured only on 2.2 
0.4 m2. The critical parameter is high ash content (1-1.2
m/m% in case of woods), therefore a special boiler is
needed. However, the ash content can be decreased by
producing pellets or briquettes, which compression
techniques are reported to change the combustion
characteristics of both ligneous and herbaceous plants
[20, 21].
4. Conclusions
The present study provided evidence that sweet
sorghum is able to adapt to the unfavorable ecological
conditions of heavy metal contaminated soils. It could
be cultivated effectively on a brownfield site after
custom soil improvement procedures. The sweet
sorghum grown on brownfield site was successfully
integrated to biorefinery in a pilot-scale system. The
total biomass was processed, namely the liquid phase
to ethanol production and the solid phase to energy
production (i.e. biogas production and combustion).
Additionally, we noticed that this sorghum variety
(Sucrosorgo) can accumulate high levels of zinc, thus
seems suitable for phytoremediation applications, too.
Based on our results sweet sorghum seems to be an
applicable plant for the economic utilization of
brownfield sites with complex processing by
biorefinery.
Acknowledgments
The authors are grateful for the financial support from
NKTH (National Office for Research and Technology,
Hungary) by the grants NKFP_07_4-BIOFINOM and
TECH_08_A4_MOLTVKBA.
References
[1] L. Oliver, U. Ferber, D. Grimski, K. Millar, P. Nathanail,
97
N S Combustion heat Calorimetric value
(m/m%) (m/m%) (kJ/kg) (kJ/kg)
0.82 0.03 15269 13714
The scale and nature of European brownfields, in: Fifth
Framework Programme, 2005.
[2] E. Karen, B.R. Glick, B.M. Greenberg, Phytoremediation
and rhizoremediation of organic soil contaminants:
Potential and challenges, Plant Science 176 (2009)
20-30.
[3] C. Garbisu, I. Alkorta, Phytoextraction: A cost-effective
plant-based technology for the removal of metals from the
environment, Bioresource Technology 77 (2001) 229-236.
[4] M.I. Lone, Z. He, P.J. Stoffella, X. Yang,
Phytoremediation of heavy metal polluted soils and water:
Progresses and perspectives, Journal of Zhejiang
University Science B 9 (2008) 210-220.
[5] H. Ohara, Biorefinery, Appl. Microbiol. Biotechnol. 62
(2003) 474-477.
[6] P. Kaparaju, M. Serrano, A.B. Thomsen, P. Kongjan, I.
Angelidaki, Bioethanol, biohydrogen and biogas
production from wheat straw in a biorefinery concept,
Bioresource Technology 100 (2009) 2562-2568.
[7] E. De Jong, R. Van Ree, R. Van Tuil, W. Elbersen,
Biorefineries for the chemical industry-A Dutch point of
view, in: B. Kamm, M. Kamm, P. Gruber (Eds.),
Biorefineries-Biobased Industrial Processes and Products,
Status Quo and Future Directions, WILEY-VCH,
Weinheim, 2005.
[8] S. Octave, D. Thomas, Biorefinery: Toward an industrial
metabolism, Biochimie 91 (2009) 659-664.
[9] G. Máthéné Gáspár, A. Anton, Mitigation of the harmful
effects of toxic element pollution by phytoremediation,
Agrokémia és Talajtan 53 (2004) 413-432. (in Hungarian)
[10] A. Murányi, L. Ködöböcz, Heavy metal uptake by plants
in different phytoremediation treatments, Cereal Res.
Commun. 36 (2008) 387-390.
[11] J.R. Peralta-Videaa, M.L. Lopez, M. Narayan, G. Saupe, J.
Gardea-Torresdey, The biochemistry of environmental
heavy metal uptake by plants: Implications for the food
chain, The International Journal of Biochemistry & Cell
Biology 41 (2009) 1665-1677.
[12] J.M. Murillo, T. Maranon, F. Cabrera, R. Lopez,
Accumulation of heavy metals in sunflower and sorghum
plants affected by the Guadiamar spill, The Science of the
Total Environment 242 (1999) 281-292.
[13] X. Zhan, D. Wang, M.R. Tuinstra, S. Bean, P.A. Seib, X.S.
Sun, Ethanol and lactic acid production as affected by
sorghum genotype and location, Industrial Crops and
Products 18 (2003) 245-255.
[14] F. Achmad, K. Yamane, S. Quan, T. Kokugan, Synthesis
98 Possibilities of Biorefinery by Sweet Sorghum (Sucrosorgo) on Brownfield Site

of polylactic acid by direct polycondensation under [Online],

vacuum without catalysts, solvents and initiators, http://www.kvvm.hu/szakmai/karmentes/jogszab/jogszab
Chemical Engineering Journal 151 (2009) 342-350. 02/tartalom2.htm. (in Hungarian)
[15] Á. Dergez, D. Bordás, Biorefinery as a real solution for [19] H. Kovács, V. Kőrösi, K. Szemmelveisz, Comparison of
reutilization of brownfield sites, in: 3rd Biofinom the combustion characteristics and gaseous air pollutions
Workshop, Gyöngyösoroszi, 2010. (in Hungarian) of five ligneous species origined from heavy metal
[16] J. Stickel, Digital data collecting and processing system contaminated site of project “BIOFINOM”,
for a pilot bioethanol production, in: 3rd Biofinom Környezetvédelmi füzetek Biofinom, 2010, pp. 67-82. (in
Workshop, Gyöngyösoroszi, 2010. (in Hungarian) Hungarian)
[17] G. Forrai, A pilot technology for biorefinery, in: 3rd [20] C.S. Chou, S.H. Lin, W.C. Lu, Preparation and
Biofinom Workshop, Gyöngyösoroszi, 2010. (in characterization of solid biomass fuel made from rice
Hungarian) straw and rice bran, Fuel Processing Technology 90 (2009)
[18] 10/2000. (VI. 2), KöM-EüM-FVM-KHVM Collective 980-987.
Decree about the Limit Values of the Qualitative Control [21] F.F. Felfli, C.A. Luengo, J.A. Suárez, P.A. Beatón, Wood
of Underground Water and Geological Medium (10/2000. briquette torrefaction, Energy for Sustainable
(VI. 2), KöM-EüM-FVM-KHVM Együttes Rendelet, Development 9 (2005) 19-22.
Journal of Environmental Science and Engineering A 1 (2012) 99-107
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Biofertilizer Utilization to Increase P Availability on Acid

Soil and Calculation of Phosphate Use Efficiency on
Corn by Radioisotope Technique

Tri Candra Setiawati

Department of Soil Science, Faculty of Agriculture, Jember University, East Java 68121, Indonesia

Received: May 27, 2011 / Accepted: July 21, 2011 / Published: January 20, 2012.

Abstract: Phosphorus deficiency is one of the most growth-limiting factors in acid soils. Phosphorus fixation results in low P use
efficiency in acid soils. Using isotopes as tracers, the percentage of utilization by plants of phosphate derived from many sources can be
determined. The objectives of this experiment are to evaluate the increase of soil phosphate availability in Oxisol and to calculate
phosphate use efficiency on corn affected by biofertilizer application. This research was conducted with bio fertilizer application made
from soybean (SB) biomass enriching with phosphate-solubilizing bacteria (PSB) compared to single phosphate fertilizer. 32P-free
carrier solution was applied to 12 pots with another set of 12 pots identical in all aspects but without 32P labeling. Corn was grown on an
Oxisol from Depok, West Java, Indonesia for 45 days. The 32P activity in corn was measured using Liquid Scintillation Counter (LSC),
while the total P-uptake of corn was analyzed using a spectrophotometer. The result showed that application of bio fertilizer has
increased soil phosphate availability in Oxisol, concentration of P on shoot and root, and phosphate uptake by corn. P-fertilizer use
efficiency (PUE) of maize was very low and less than 1% up to 45-days after planting. Radioisotope technique by the use of 32P in this
research was not effective yet to trace the P transport from soil to plant.

Key words: Biomass, P-use efficiency (PUE), phosphorus percentage derived from fertilizer (%Pdff), phosphorus percentage derived
from soil (%Pdfs), 32P labeling.

1. Introduction which is in the edge of clay mineral, and

oxide-hydroxide Al and Fe both in the form of
Most of phosphate applied to the acid mineral soil
amorphous and crystalline. In addition, the kaolinite
through P inorganic fertilizer in a large and continuous
mineral which dominates Oxisols also takes part in P
amount is fixated quickly into insoluble form by clay
adsorbtion. It’s mechanism occurs due to both of Al
minerals and the ions of Al, Fe, and Mn which are
and Fe ions are on clay surface, associated ion of P with
soluble. Most of them are influenced by soil reactions,
OH in edges of kaolinite, and P ions enter to the surface
thus making it not available for the plants. Phosphate
of amorphous clay.
deficiency on the acid soil is mainly caused by
Phosphorus sorption may decrease with increasing
adsorption of H2PO4- by Al and Fe hydro (oxide),
pH in acid soils which is caused by precipitation of
which causes most of total-P in the soil in unavailable
amorphous Fe and Al oxides, a greater competition of
forms for plants. Generally, the soil P will form AlPO4
OH- with phosphate ions for adsorption and an increase
and FePO4 which are difficult to dissolve [1]. Other
of negative charges on soil particles [2]. Organic
fractions which also fixate the large amount of P
anions with low molecular mass may form or make
(adsorption P) are polymeric of Al and Fe, Al ion
stable complexes with soluble Fe or Al that prevent
Corresponding author: Tri Candra Setiawati, Ph.D., main formation of Fe or Al phosphates [3, 4], and also with
research field: soil biology. E-mail: allophane minerals [5] of which this mechanism is
candra_setiawati@yahoo.com.
100 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use
Efficiency on Corn by Radioisotope Technique
caused by strong affinity of organic anions. Organic
acids may also reduce P adsorption by Fe or Al
hydroxides due to the competition with phosphates and
blockage of the adsorption sites.
The amount of phosphate from inorganic fertiliser
that can readily uptaken by plant depends on soil type
and plant. Only about 10%-20% of phosphorus are
used by crops during the season [6]. Organic fertilizer
which is enriched with phosphate solubilizing
microorganism is used as “miners agent” of phosphate
residue in the soil because it has some advantages such
as: energy-save, unpollute material, improve chemical
and biological as well as physical characteristics and
able to dissolve P become soluble, prevent P from
fertilizer by adsorption elements and reduce toxicity
Al3+, Fe3+, and Mn+3 towards plants in the acid soil.
Phosphate-solubilizing microbes improve
availability of P from insoluble P on several types of
soil effectively [7-9]. In addition, it also has abilities to
fractionation of P forms in the acid soil [8]. If the
concentration of P in soil solution is high enough, then
the movement of phosphate ion is generally caused by
the diffusion process. The mass flow process may have
role in that transportation. Soil with concentration P of
1 ppm , the mass flow has 20% contribution from those,
which is uptake by plants [10].
The ion on the root surface will go to the outer cavity
of root through a simple diffusion process, the
adsorption change, and the activities of carrier material.
Then, the ion passes a cavity on the root by using
metabolism energy that is called as active adsorption
[11]. The active uptake of P by plants directly moves
through the root. There are two kinds of transporters
molecules which manage the plants P uptake i.e.: (1)
transporters molecule with high affinity transporters,
that act if the P concentration in the soil solution is low;
and (2) the transporters molecule with low affinity
transporters that work without referring to P
concentration in the soil solution. Transporters
molecule is on the cell membrane and flow phosphate
ion from the outside to be transported to the cell. This
process will decrease pH of cell cytoplasm. To balance
it, the plants release hydrogen ion through proton pump
which also happens in the cell membrane [12].
Many isotopic techniques can be applied to
determine the relative immediate and residual
effectiveness of P fertilizers. Using isotopes as tracers,
the percentage of utilization by plants of the P derived
from a fertilizer can be determined. Labeling of plant
residues with the radioisotopes 32P or 33P has been used
to investigate the recovery of P from plant residues in
soil P pools and in growing plants. The radioisotopes
32
P can also be added to soil without simultaneous
application of 31P (i.e., carrier free). In this case, the
amount of P introduced with the radioisotope is
negligible in relation to native P in the to soil solution
and other soil P pools, but the the radioisotope can
easily be detected because of the sensitivity of β
counting. Carrier-free addition of 32P or 33P is
commonly used to determine isotopically
exchangeable P in soil-solution mixtures and in pot
experiments [13, 14].
Ortho phosphoric acid (32P) uses as a fertilizer on
tomato by Mousavishalmani et al. [15]. The result
showed that only 23.1% is uptake by plant (11.7% on
fruit, 7.8% stem, 2.2% leaf and 1.6% root), and
approximatelly 76.7% is as residue in soil. Other
research is combination of triple super phosphate
(TS32P), organic matter (green manure and manure)
and CaCO3, which use as material to test the efficiency
of phosphate fertilizer on maize [16]. The research
explained that P-use efficiency (PUE) by maize was
only from 6.6% up to 7.98%.
The objectives of this research were: (1) to evaluate
increasing phosphate available on Oxisol; (2) to
increase P uptake by corn; and (3) to calculate
phosphate use efficiency (PUE) by corn from many
sources.
2. Materials and Methods
The experiment was conducted in plastic green
house located at Agriculture Faculty, Jember
 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use 101
Efficiency on Corn by Radioisotope Technique

University in Jember-East Java. Soil (Typic kandiudox)
was collected from Cipayung Jaya, Depok, West Java
(Table 1).
The experimental layout was a completely
randomized design using three replications:
 Control (C);
 Biofertiliser (PH);
 Biofertiliser + Phosphate Rock (4.5 g·pot-1)
(PHR);
 SP-36 (3 g·pot-1) (SP).
A set of 12 pots were applied with 20 mL·pot-1
carrier free solution form of KH232PO4 with total
activity of 1,415.2 µCi·pot-1 to get residue data P and P
use efficiency from the source of P, while another set of
experiment is without isotope application to get the
harvest is at 45-day after planting when maize are on
generative fase. The soil and plant tissue sample are
analyzed for some parameters i.e.: FSB population,
concentration of P-plant, P-uptake, soil P (total,
available and organic form) and pH. The 32P activity
was counted by the Carenkov method using liquid
scintillation counter analyzed at the soil laboratory,
Centre for Research and Development of Isotopes and
Radiation Technology, National Nuclear Energy
Agency (BATAN).
From the radio assay data, the %Pdff, %Pdfs, and P
use efficiency were computed by Eqs. (1)-(5):
% Pdff
SA of 32P plant treated with test P (fertilizer or microbe)
=1 ×100 (1)
SA of 32P plant not treated with test P

data of viability phosphate solubilizing bacteria and dpm·g-1 plant

availability P soil so that there were 24 experiment pots. Specific activity = (2)
mg P·g-1 plant
The com seed was given to 4 plants in each pot and was
thinned out to 2 plants each pot after one week. The %Pdfs = 100 %Pdff (3)
base fertilizer was given 2.7 g urea for each (given Uptake from fertilizer (mg P·pot-1)
-1
three times) and 2.04 g of KCl for 7 kg·pot at the %Pdff
= × Total P-uptake (mg P·pot-1) (4)
planting period. 100
Soybean biomass will be enrichment with phosphate
P-use efficiency (%PUE)
solubilizing bacteria (PSB) (Bacillus sp) used as P uptake from fertilizer (mg P·pot-1)
biofertilizer. The dosage is 80 g·pot-1 with PSB density =- × 100(5)
Fertilizer P applied (mg P·pot-1)
is ± 5 × 109 cfu. It applies at the planting time. Plant
Table 1 Initial Typic kandiudox characteristic.
Parameter Value Parameter Value
pH: H2O 4.3 CEC (NH4-Acetic 1 N pH 7) (cmol(+)·kg-1) 16.64
CEC clay (NH4-Acetat 1 N pH 7)
KCl 3.8 27.73
(cmol(+)·kg-1)
Organic-C (%) 1.82 K2O extract HCL 25% (mg·100g-1) 7
Total-N (%) 0.19 Al3+ (KCl, 1N) (me·100 g-1) 2.94
C/N 10 Al saturation 53.07
P2O5 (mg·100 g-1)
41 Al (dithionite) 1.77
-Extract HCl 25%
-Total-P (ppm) 257 Al (Oxalate) 0.64
-Bray I (ppm) 3.5 Al (Phyrophosphate) 0.21
P-Retension (%) 65.5 Fe (dithionite) 5.20
CaCO3 (mg·kg-1) 765 Fe (Oxalate) 2.00
Exchangeable base (NH4-Acetic 1 N pH 7) (cmol(+)·kg-1) Fe (Phyrophosphate) 0.21
Ca 1.56 Texture clay
Mg 0.56 Sand 2
K 0.09 Silt 38
Na 0.15 Clay 60
102 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use
Efficiency on Corn by Radioisotope Technique
3. Result and Discussion
3.1 The Influence of Biofertilizer towards PSB
Population and Concentration of P on Oxisol
The change of population of phosphate solubilizing
bacteria in the soil is influenced by several factors: the
content of organic mater, nutrition availability,
environment situation, the presence of competitor
organism and other factors.
3.2 Total Population of Phosphate Solubilizing
Bacteria (PSB)
Total populations of phosphate solubilizing bacteria
in the soil are the total PSB indigenous and inoculants
on the biofertilizer. At the beginning to H-15 there is a
decrease of population on all treatments. The model of
PSB grows on the treatment by adding biofertilizer and
a natural phosphate (PHR) was different from other
treatment (Fig. 1a). The population decreases to H-30
and it increases in H-45. PSB population on the
treatment by adding biofertilizer (PH and PHR) is
better compared to the control and SP-36 treatment,
which shows that a better environment can be presence
of biofertilizer.
Generally, PSB is heterotrophy microorganism. The
addition of organic carbon as the energy source from
the organic matters decompositions will increase the
PSB proliferation, increase nutrient, carbon and
(a)
Fig. 1 The influence of treatment towards (a) PSB proliferation and (b) concentration of P available in the soil to H45.
improve the soil chemical characteristic. Besides,
during the plants growth, the root both actively and
passively releases exudates in organic form mainly in
the form of carbohydrate, carboxylic acid and amino
acids which are useful as the source of carbon, nitrogen
and energy to accelerate microbial growth in
rhizosphere. In the research by Baudoin et al [17], the
population of bacteria on the rhizosphere of corn is
influenced by concentration of carbon which is related
to C/N ratios.
3.3 Concentration of Soil P-Available
The application of biofertilizer (PH) which is
combined both with RP and without RP increases the
P-available in the soil to H-45 compared to the control
treatment and SP-36 (Fig. 1b). Input of plants biomass
rich with PSB as biofertilizer can enhance P in the
form of organic P, which can be mineralized
enzymatically by several mechanisms [18] to an
organic P which can be used by the plants. Bacillus sp.
as the phosphate solubilizing bacteria (PSB) in the
biofertilizer produce alkaline phosphatase enzyme as
well as organic acid such as citrate acid, acetate etc. as
the result of secondary metabolism [19]. Most of the
PSB genus such as Bacillus Pseudomonas, and
Azotobacter is known to produce several organic acids
[4]. The organic acids as the secondary metabolite of
microbe and the result of root secretion increase P soil
(b)
 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use
Efficiency on Corn by Radioisotope Technique
availability through several mechanisms: (1) P release
which is bound by Fe, Al and Ca so that there is a
complex forming of organic metals [20]. It causes
movement of P anion to the root surface. If the ratio of
[Fe/Al] towards [P] > 1, there will be ligand exchange
mechanism [21]; (2) citric ion or other ions (such as
malate, acetate, and oxalic) can be absorbed by
adsorption site which is the same as phosphate anion so
that there is P release simultaneously; (3) organic acid
adsorption (citric acid, etc) changes the surface charge
to become negative charge. Besides, the presence of
carbon addition gained from biofertilizer (soybean
biomass) can reduce the phosphate diffusion to the soil
pores which can adsorb P [22].
In addition, during the decomposition process, the
plants biomass stimulate P availability by releasing
CO2 in the form of H2CO3 or organic acid which is
released in the soil solution, resulting mineral
solubilise containing of P. On the soil with high
capacity of P fixation, organic acid blocked the
adsorption site of P or through anion exchange [23]. On
the contrary, P availability does not increase when
there is P-available competition between plants with
the microorganisms. Concentration of P availability
reduces because microorganims use phosphate to form
the microbial biomass of P and the process of
re-adsorption in the soil. This mechanism causes the
decrease of P-concentration on H-30 to H-45, besides
(a)
Fig. 2 The influence of treatment towards (a) P concentration on tissue and (b) the P uptake by plants on H-45.
103
an adsorption by plants. The results of the research
done by Guppy et al. [24] show that the addition of
soybean extract on the Oxisol only block for a while of
P adsorption, and after 6 days there is no blocking
competition, so that there is no increase in P
availability. According to Dye [25], the effectiveness
of organic acid addition in reducing adsorption P
depends on the soil pH and organic molecule structure.
3.4 The Influence of Biofertilizer Application towards
P-Tissue Concentration and P Plants Adsorption
The application of biofertilizer and RP (PHR) is
bigger than control and the SP-36 treatment on P
concentration in the plant tissue on shoot and root and
P adsorption on H-45 (Fig. 2).
P concentration of shoot is between 0.92 to 2.24 mg
P·g-1 and included as low category, while on the root is
0.45 to 1.30 mg P·g-1. In this research there is a
correlation between soil P-available with P
concentration in the plant tissue (root and shoot) with
correlation coefficient value R2 = 0.78. This value
shows that the root is effective enough to adsorb P soil
and translocation to the other part of the tissue. This
effectively can be seen in the high correlation
between concentration of P-tissue with P adsorption
with R2 = 0.879 (shoot) and R2 = 0.888 (root). The
roots uptake P and effectively translocation to the parts
of the plant and increase the biomass, so that the plants
(b)
104 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use
Efficiency on Corn by Radioisotope Technique
adsorption is high. The plants develop a number of
morphological, physiological, biochemical and
molecule responses to solve the low P-inorganic
condition including morphological change and root
architecture, the accumulation of anthocyanin pigment,
phosphomonoesterase enzyme and organic acids
secretion to rhizosphere to increase the efficiency in
taking P-inorganic and metabolism change [12].
According to Rausch and Bucher [12], the presence
of P-inorganic (Pi) being adsorbed to the root
symplasm has five possibilities: (1) enter to the place of
metabolism (metabolic pool) that is cell cytoplasm
with its organelle; (2) a small part of it enters the way
of P lipid biosynthetic, DNA, and RNA which then
become the cell component structure; (3) several of
them loss from the cell through efflux, especially on the
high Pi condition; (4) Pi enters and is stored in the
vacuole to regulate the Pi presence which is balance in
the cell (homeostatic); and (5) Pi is transported through
symplast towards parenchyma cells of xylem and then
is released to appoplast to be translocated to the top
part of the plants.
The biofertilizer application which is combined with
RP (PHR) contributes P on the soil continuously
because the decomposition process of soybean biomass
and Bacillus sp activities. On that treatment, P
contributes in the form of P-organic and P-inorganic so
that there is significant difference compared to other
treatment and control.
3.5 The Influence of P-Use Efficiency (PUE) (%)
Treatment by the Plants and P Plants Adsorption from
Each Source P (%Pdft)
The treatments (PH, PHR and SP) increase the dry
weight of the plants (shoot and root) and the P
accumulation in plant tissue (Figs. 3a and 3b). PH
application with RP is the best treatment.
Based on the value of specific activities (SA) which
describe the proportion of 32P/31P on the shoot,
application of biofertilizer (PH) and the combination of
biofertilizer with RP (PHR) is more effective in
releasing P compared to the control and SP (Fig. 3c),
while on the root, the specific activities value control is
bigger than the treatment, this is because the low of
stable P (31P) which is translocated to the root.
The result shows that plants uptake P from the
biofertilizer much more than the P uptake from SP-36
(Table 2). Biofertilizer provides more P for the plants,
so that the uptake of P is more than other source. This
can be seen from the value of % Pdff PH > SP. This
phenomenon shows P release from the biofertilizer
occurs continuously and uptake efficiently by the root.
Biofertilizer is enriched with the
phosphate-solubilizing bacteria, which can release P
with several mechanisms, i.e. P organic mineralization
with several is due to phosphates enzyme, the blocking
of adsorption site or through anion exchange [23].
The P-use efficiency from fertilizer (PUE) to H-45 is
still low that is maximal 0.598% on the PHR treatment
(Table 2), also P uptake from each fertilizer is low. It
means that the plants uptake more anion phosphate
from soil, this condition caused by the result the
fertilizer residue to the following period is still high.
Several possibilities of mechanisms which happen are
re-adsorption by the soil adsorption agent and
re-binding by the metals in the soil solution. Both of
mechanisms in Typic kandiudox occur continuously.
This is also seen from a high P retention (> 65%). In
addition, based on the Oxisol characteristics which is
used with ratio Feo/Fed (Fe oxalate extract/Fe
dithionate extract) = 0.38, it shows that Oxisol is the
mature soil which is indicated by ratio Feo/Fed < 0.75.
The major forms are free oxides of Al and Fe as well as
active Al and Fe. This is proved by that the Al and Fe
concentration with dithionite extract is high (1.77%
and 5.20%); Al and Fe (oxalate extracted) is 0.64% and
2.00%. One of mechanisms that happens is P
adsorption or re-binding to Fe oxide. Re-adsorption
usually has two steps: initial adsorption which happens
quickly on the outer surface and followed by slow
adsorption which may happen for days or weeks. Slow
adsorption of P occurs through diffusion to crystal
 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use 105
Efficiency on Corn by Radioisotope Technique

(a)

(b)

(c)
Fig. 3 (a) Plant dry weight; (b) concentration of P tissue; (c) specific activities .

Table 2 The influence of treatment %Pdff, %Pdfs, P-uptake from the fertilizer PUE (%).

%Pdff %Pdfs P-uptake from fertilizer (mg P/pot) PUE (%)

Treatment
Shoot Root Shoot Root Shoot Root Shoot Root
+ PH 2.408 0.768 97.592 99.232 0.177 0.006 0.443 0.016
+ PH + RP 1.688 0.602 98.312 99.398 0.239 0.008 0.598 0.020
+ SP-36 0.669 0.458 99.331 99.542 0.088 0.007 0.219 0.017
Pdff = P from the fertilizer; Pdfs = P from the soil; PUE = use efficiency of P by plants.
106 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use
Efficiency on Corn by Radioisotope Technique

micro porous, an area between micro-meso porous [9] T.C. Setiawati, Viability and Capability Test of Phosphate
Solubilizing Bacteria on Combination of Humic and
crystal and aggregate particle [22].
Zeolite Medium to Dissolve Soil Phosphate, Fundamental
research report, Faculty of Agriculture, University of
4. Conclusion
Jember, 2003.
Application of biofertilizer can increase P [10] S.A. Tisdale, W.L. Nelson, J.M. Beaton, J.L. Havlin, Soil
Fertility and Fertilizers, Macmillan Publishing Co., New
availability on the Oxisol, P concentration on the tissue
York, 1993.
(shoot and root) and the P uptake by plants. The P use [11] M.E. Premono, Phosphate solubilizing microorganism:
efficiency by the plants which comes from biofertilizer Influence on soil phosphate and efficiency of phosphate
or SP-36 (PUE) is still low (1%) so that PH and SP fertilizer on sugarcane, Doctor Dissertation, Bogor
Institute of Agriculture, 1994.
application remain a high content of residue. The use of
[12] C. Rausch, M. Bucher, Molecular mechanisms of
radioisotope 32P in the research is not effective enough phosphate transport in plants, Planta 216 (2002) 23-37.
to trace P transport from plants to soil. [13] J.C. Fardeau, G. Guiraud, C. Marol, The role of isotopic
techniques on the evaluation of the agronomic
References effectiveness of P fertilisersm, Fertiliser Research 45
(1996) 101-109.
[1] J.O. Agbenin, Extractable iron and aluminum effects on
[14] E.K. Bunemann, F. Steinebrunner, P.C. Smithson, E.
phosphate sorption in a savanna Alfisol, Soil Sci. Soc. Am.
Frossard, A. Oberson, Phosphorus dynamics in a highly
J. 67 (2003) 589-595.
weathered soil as revealed by isotopic labeling techniques,
[2] R. Naidu, J.K. Syers, R.W. Tillman, J.H. Kirkman, Effect
Soil Sci. Soc. Am. J. 68 (2004) 1645-1655.
of liming and added phosphate on charge characteristics of
[15] M.A. Mousavishalmani, N. Sagheb, M.S. Hobbi, H. Rafh,
acid soils, Journal of Soil Sci. 41 (1990) 157-164.
A. Khorasani, Fertilizer P distribution into different parts
[3] M. Pigna, A. Violante, M. Ricciardella, Adsorption of
of plant and soil under trickle fertigation on tomato by 32P,
phosphate and sulphate on metal oxides and variable
in: 17th Word Congress Soil Science, Paper 2286,
charge soils as affected by organic ligands, in: 17th
Thailand, 2002.
Thailand Symposium, Paper 134, 2002.
[16] N. Hakim, Organic matter for increasing P-fertilizer use
[4] S. Srivastava, M.T. Kausalya, G. Archana, O.P. Rupela, G.
efficiency of maize in Ultisol by using 32P technique, in:
Naresh-Kumar, Efficacy of organic acid secreting bacteria
17th Thailand Symposium, Paper 229, 2002.
in solubilization of rock phosphate in acidic Alfisols, in:
[17] E. Baudoin, E. Benizri, A. Guckert, Impact of artificial
First International Meeting on Microbial Phosphate
root exudates on the bacterial community structure in bulk
Solubilization, 2007, pp. 117-124.
soil and maize rhizosphere, Soil Biology and Biochemistry
[5] A. Violante, L. Gianfreda, Role of biomolecules in the
35 (2003) 1183-1192.
formation of variable charge minerals and organo-mineral
[18] B.B. Jana, Distribution pattern and role of phosphate
complexes and their reactivity with plant nutrients and
solubilizing bacteria in the enhancement of fertilizer
organic in soil, in: J.B. Bollag, G. Stotzky (Eds.), Soil
value of rock phosphate in aquaculture ponds:
Biochemistry, Marcell Dekker, New York, 2000, pp.
State-of-the-art, in: First International Meeting on
207-270.
Microbial Phosphate Solubilization, Springer, 2007, pp.
[6] L.L. Hammond, R.B. Diamond, Effectiveness of
229-238.
alternative phosphate fertilizer in tropical agriculture,
[19] T.C. Setiawati, P. Miharja, Identification and
International Fertilizer Development Center (IFDC),
quantification of phosphate solubilizing bacteria
Muscle Shoal Alabama, 1987.
metabolites and their effect to Rhizoctonia solani activity
[7] T.C. Setiawati, Effect of phosphate-solubilizing microbe
in soybean, Jurnal Tanah Tropika, 2008.
to improve phosphate availability and growth of tobacco
[20] J. Gerke, Phosphate, aluminium and iron in the soil
Besuki Na-Ogst (Nicotiana tabacum. L), Thesis, Bogor
solution of three different soils in relation to varying
Institute of Agriculture, 1998.
concentrations of citric acid, Journal of Plant Nutrition
[8] T.C. Setiawati, Viability and Effectively of
and Soil Science 155 (1992) 339-343.
Phosphate-Solubilizing Microbe on Liquid Medium to
[21] K.D. Earl, J.K. Syers, J.R. Mclaughlin, Origin of the
Improve Phosphate Availability, Matching-Grand Asian
effects of citrate, tartrate, and acetate on phosphate
Development Bank (ADB) research report, Faculty of
sorption by soils and synthetic gels, Soil Sci. Soc. Am. J.
Agriculture, University of Jember, 2000.
43 (1979) 674-678.
 Biofertilizer Utilization to Increase P Availability on Acid Soil and Calculation of Phosphate Use 107
Efficiency on Corn by Radioisotope Technique
[22] C. Mikutta, G. Neumann, F. Lang, Phosphate desorption [24] C.N. Guppy, N.W. Menzies, F.P.C. Blamey, P.W. Moody,
from goethite in the presence of galacturonate, Do decomposing organic matter residues reduce
polygalacturonate, and maize mucigel (Zea mays L.), Soil phosphorus sorption in highly weathered soils?, Soil Sci.
Sci. Soc. Am. J. 70 (2006) 1731-1740. Soc. Am. J. 69 (2005) 1405-1411.
[23] M.A. Cavigelli, S.J. Thien, Phosphorus bioavailability [25] C. Dye, Effect of citrate and tartrate on phosphate
following incorporation of green manure crops, Soil Sci. adsorption by amorphous ferric hydroxide, Fertilizer
Soc. Am. J. 67 (2003) 1186-1194. Research 40 (1995) 129-134.
Journal of Environmental Science and Engineering A 1 (2012) 108-114
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Anaerobic Biodegradability of Selected Biodegradable

Plastics and Biobased Products

Hamed M. El-Mashad1, 2, 3, Ruihong Zhang1 and Joseph P. Greene4

1. Biological and Agricultural Engineering Department, University of California Davis, CA 95616, USA
2. Department of Agricultural Engineering, Faculty of Agriculture, Mansoura University, El-Mansoura 35516, Egypt
3. Department of Agriculture and Environmental Sciences, Lincoln University, Jefferson City, MO 65101, USA
4. Department of Mechanical Engineering, Mechatronic Engineering, and Manufacturing Technology, California State University,
Chico, CA 95929, USA

Received: November 2, 2010 / Accepted: August 2, 2011 / Published: January 20, 2012.

Abstract: This study was conducted to investigate the biodegradability of six biodegradable plastic products and two cellulose-based
products that are commercially available in the United States. The six biodegradable plastics included four films, one straw and one cup.
The four plastic films (namely first, second, third and fourth film) were made of the following materials, respectively: corn starch plus
polycaprolactone, polyhydroxyalkanoates (hydroxy fatty acids), poly tetramethylene adipate-co-terephthalate, and polyethylene plus
prodegradant additives. The straw and cup were made of polylactide. In addition to these products, one synthetic plastic film was tested
as a comparison. The two cellulose-based products included sugar cane plates and Kraft paper. The anaerobic biodegradability of each
product was determined by using batch anaerobic digestion tests at a thermophilic temperature of 50 ± 2 oC with the food waste used as
the source of nutrients needed for microbial growth. The results showed that after about 43 days of digestion, the biogas yields were
0.836, 0.187, 0.104, 0.061, 0.053, and 0.024 L/gVS of second film, first film, fourth film, plastic, third film, and straws respectively.
Negligible biogas was produced from plastic cup. The digestion test results indicated that the degradation of biodegradable plastics
started after the depletion of the biodegradable fraction of food waste. The results of this research are useful for assessing the fate of
various products in anaerobic digesters treating municipal solid wastes containing non-recycled biodegradable plastics.

Key words: Biodegradable plastics, biogas, anaerobic digestion.

1. Introduction environmental problems with many public works [3].

This is due to the high volume to weight ratio of
Plastics are very important for many domestic,
plastics and their resistance to natural degradation in
agricultural and industrial applications. The annual
landfills or waste treatment facilities. Moreover,
world production of synthetic polymers is about 140
synthetic polymers represent a large demand of fossil
million tons [1]. Plastics represent about 12% and 25%,
fuel. In comparison, biodegradable plastics and
respectively of the total weight and volume of
biobased materials provide environmentally friendly
municipal solid waste (MSW) generated in the United
and sustainable alternatives for consumers.
States [2]. At present, only about 5% (by weight) of the
Biodegradable plastic may be defined as a plastic
plastic is recycled. The vast increase in plastics
derived mainly from renewable raw materials [4]. The
consumption has led to the so-called “white pollution”,
biodegradability of polymers depends on the physical
taking up large volume in landfills and causing
properties and chemical characteristics of polymers as
well as the environmental conditions [5, 6]. The
Corresponding author: Ruihong Zhang, Ph.D., professor, physical properties could include particle size, density,
research fields: biomass processing and bioenvironmental
engineering. E-mail: rhzhang@ucdavis.edu. and porosity. The chemical characteristics could be
 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products
accessibility of the polymer to enzymes, molecular
weight distribution of the polymer, presence of
non-biodegradable polymers, chemical linkage in the
polymer backbone, the position and the chemical
activity of pendant groups, and the type and the
chemical activity of end-groups [7]. The environmental
conditions could include moisture and oxygen contents,
temperature, and types and concentrations
microorganisms. Currently the definitions
standards for biodegradable plastics are all based on
their biodegradability in a composting process [3].
Though anaerobic digestion processes become
attractive methods for treating municipal solid waste
due to their potential for bioenergy recovery, there is a
need for better understanding of the fate of
biodegradable plastics and biobased polymers under
anaerobic conditions. Ishigaki et al. [8] reported a study
on the degradability of four kinds of commercial
biodegradable plastics under aerobic and anaerobic
conditions. The products were polyhydroxybutyrate
and hydroxyvalerate plastic, polycaprolactone plastic,
blend of starch and polyvinyl alcohol plastic and
cellulose acetate plastic. Their results showed that after
120 days, the biodegradable plastics derived from
natural polymers, such as starch or cellulose, contained
recalcitrant components to microbial degradation;
anaerobic degradation behavior of commercial
biodegradable plastics was different from pure
polymers due to the additives used to improve the
performance of the final product.
Polyhydroxyalkanoates (PHAs) was produced from
municipal waste. Lee and Yu [9] examined the
biodegradability of the PHAs using aerobic and
anaerobic degradation experiments conducted at 20 °C.
The weight loss of the plastic film was measured over
time. Their results showed that PHAs were degraded
quickly: about 70% and 75% of the mass was reduced
after six weeks under anaerobic and aerobic conditions,
respectively. They also found that the degradation of
PHAs was very low (about 20%) in soil at the same
temperature and time period.
109
Polycaprolactone was not easily anaerobically
degraded under mesophilic conditions (35 oC) [10].
While under thermophilic conditions (55 oC), poly
lactic acid and polycaprolactone were highly
biodegradable [11]. In light of the advantages
associated with thermophilic digestion for waste
treatment, such as faster digestion rate and higher
of pathogen destruction capability, as well as higher
and instability of the plastics at a higher temperature [12],
the thermophilic temperature was used in the batch
anaerobic digestion experiments in this study. The
objective was to determine the biodegradabilities of
six selected biodegradable plastic and two
cellulose-based products in terms of their biogas
production potential. The results from this study will be
very useful for the selection of plastic or
polymer-based consumer products as well as their
treatment and disposal conditions in anaerobic
digestion systems designed for treating municipal solid
wastes.
2. Materials and Methods
2.1 Materials
Six biodegradable plastics and two cellulose-based
products were tested in this study. The six
biodegradable plastics included four films, one straw
and one cup. The four plastic films (namely first,
second, third and fourth film) were made of the
following materials, respectively: corn starch plus
polycaprolactone, polyhydroxyalkanoates (hydroxy
fatty acids), poly tetramethylene
adipate-co-terephthalate, and polyethylene plus
prodegradant additives. The four films are used mainly
for bags production. The straw and cup were made of
polylactide. One synthetic plastic film was also tested
as a comparison. The two cellulose-based products
included sugar cane plates and Kraft paper. The sugar
cane plates are used as dining plates. The compositions
of tested products are shown in Table 1.
These products were codigested with food waste for
two reasons. Firstly, the non-recycled plastics
110 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products
generally are present in MSW that contains food waste
and secondly, food waste can serve as a source for
nutrients necessary for microbial growth. Each product
was cut manually into about 0.5 cm squares. Food
waste was provided by a waste management company
in Northern California. Zhang et al. [13] presented
details about the sources and processing procedures of
food waste.
2.2 Batch Anaerobic Digestion Experiments
The experiments were carried out in anaerobic batch
reactors with 1,000 mL total volume in each reactor.
The digestion tests of biodegradable plastics were
performed at an initial substrate loading of 4 gVS/L,
which contained 2 gVS/L of food waste and 2 gVS/L
from each tested product. The F/M ratio (VS of
feedstock/VS of inoculum) was about 2. In all the
experiments, 200 mL of adapted, for each plastic type,
anaerobic inoculum were used. This adapted inoculum
was prepared by digesting the same plastics for 20 days
using the sludge collected from an anaerobic digester at
the wastewater treatment plant in Davis, CA (data are
not shown). The TS and VS contents of that inoculum
were 23 and 11.3 g/L, respectively.
Inoculum was added into each reactor at the
beginning of the experiments. Then the required
amounts of products and food waste were added to
each reactor and reactors were filled to an effective
Table 1 Composition of the products studied [15].
Category Product
Film 1
Film 2
Biodegradable plastics
Film 3
Film 4
Plastic straw
Plastic cup
Plates
Cellulose-based products
Paper
Synthetic plastics Plastic
volume of 500 mL with tap water. Finally, rubber
septum and a screw cap were placed to seal each
reactor. To assure anaerobic conditions, each reactor
was purged with helium gas for five minutes. Two
reactors contained 2 gVS/L food waste, 200 mL of
inoculum and tap water were used as a control to
correct for the biogas produced from food waste and
inoculum. All experiments were performed in
duplicate at thermophilic temperature of 50 ± 2 oC. The
average initial pH of all reactors was 7.4.
After one week, the reactors containing plates and
Kraft paper were inhibited, as noticed by very low
biogas production, as compared with others. Therefore,
the reactors of these two products were similarly
started again but by using 1 gVS/L of food waste and 2
gVS/L from each product.
2.3 Proximate Analysis
Total solid and volatile solids of the anaerobic
inoculum and food waste were measured according to
the standard methods [14]. The total solids of plastics
were also measured according to the same standard
methods. Due to the problem encountered in the VS
measurements of plastics by the APHA method [14],
the volatile solids of plastics were measured according
to D5630-06 method [16]. In this procedure samples
were flamed over a burner before ashed in a muffle
furnace at 550 oC.
Composition
Made from the Novamont resin, derived from corn starch, in
combination with fully biodegradable aliphatic polyesters,
aliphatic/aromatic polyesters or in particular polylactic acid.
Made of polyhydroxyalkanoates that are biopolyesters composed of
hydroxy fatty acids.
Is analiphatic-aromatic copolyester based on 1,4-butanediol and the
dicarbonic acids, adipic acid and terephthalic acid.
Is a low density polyethylene (C2H4)n plus cobalt stearate (1%-5%).
Made of polylactide that is aliphatic polyester (C68H108O8Ti2).
The same as plastic straw.
Made of sugar cane bagasse that is fibrous material left over after
pressing out sugar cane juice.
Kraft paper, cellulose based.
Made of low density polyethylene (C2H4)n.
 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products 111

2.4 Biogas Measurements left after burning the sample at 550 oC.

The daily biogas production from each reactor was
measured. Each day, pressure was measured in the
headspace of each batch reactor using a
WAL-BMP-Test system pressure gauge. The biogas in
the reactor was then released under water to prevent
any gas exchange between the reactor and the air.
Pressure was then measured again to provide the initial
conditions for the next day’s test. Biogas volumes of
each reactor were calculated as described by
El-Mashad and Zhang [17] as follows:
P  Vhead  C
VBiogas  (1)
R T
where:
VBiogas = daily biogas volume (L);
ΔP = absolute pressure difference (mbar);
Vhead = volume of the head space (L);
C = molar volume (22.41 L/mol);
R = universal gas constant (83.14 L·mbar/K·mol);
T = absolute temperature (K).
Methane and carbon dioxide contents of the biogas
produced in each reactor was periodically measured
using a gas chromatography (GC) (HP 5890A). A 1.8 ×
0.32 mm Alltech Carbospher column was used. Helium
was used as a carrier gas at a flow rate of 60 mL·min-1.
The temperatures of oven and thermal conductivity
detector (TCD) were 100 and 120 oC, respectively. A
gas standard with 60% methane and 40% carbon
dioxide was used for the GC calibration.
3. Results and Discussion
3.2 Biogas Production from Biodegradable Plastics
The cumulative biogas production yields and daily
biogas production rates from the digesters treating
different products and food waste are shown in Figs. 1
and 2, respectively. The data shown in these two
figures are the average of the data obtained from the
two duplicate reactors, for each product, initially
started at loading of 4 gVS/L (2 g VS of each product
and 2 gVS/L of food waste). For all the reactors, biogas
was produced immediately after the experiment started,
i.e. no lag phase was noticed. From day 3 to day 7,
there was a relatively low biogas production rate at
about 0.02 L/L·gVS·day (Fig. 2). Then biogas
production increased sharply after day 7. For food
waste reactor and all the reactors containing food waste
and the test products except for film 2, biogas
production increased till day 15, then the biogas
production became very low (Figs. 1 and 2). Biogas
production from the reactors treating film 2 and food
waste increased after day 15 till about day 35 and then
leveled off. The data shown in Fig. 1 indicate that
except for the reactors containing film 2, there was
little difference between all the products tested. As
stated above, reactors containing food waste at an
initial loading of 2 gVS/L were also tested as controls
to correct for the biogas produced from food waste and
inoculum. The biogas yields of each product was as
Table 2 Characteristics of the products and food waste.
Numbers in parentheses are standard deviations.

3.1 Characteristics of Materials Product type TS (%) VS/TS (%)

Film 1 93.48 (0.21) 99.58 (0.12)
The TS and VS contents of different products are Film 2 99.03 (0.09) 100 (0.02)
shown in Table 2. The higher moisture contents were Film 3 99.96 (0.01) 90.57 (0.48)
Film 4 100 (0.27) 96.18 (0.03)
measured to be about 5.8% and 6.5%, respectively for
Plastic 97.75 (3.13) 99.91 (0.23)
film 1 and plates products. Most of the biodegradable Cup 99.60 (0.01) 99.98 (0.02)
plastics showed very low ash contents (i.e. VS/TS were Straw 99.59 (0.01) 94.90 (0.02)
higher than 90%). The film 4 did not show any loss of Plates 94.21 (0.14) 99.43 (0.01)
moisture content after being heated at 105 oC for 24 Paper 96.64 (0.07) 95.72 (0.03)
Food waste 19.17 (0.42) 92.83 (0.12)
hours. Film 2 was 100% organic as there was no ash
112 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products
Fig. 1 Biogas yields of different biodegradable plastic and
cellulose-based products, and food waste.
Fig. 2 Biogas production rates from different plastic
products and food waste.
Fig. 3 Biogas yields of different products after 43 days of
digestion time. Y error bars are the standard deviations
between the duplicate reactors.
calculated by subtracting the cumulative biogas
production, after 43 days, from the reactors treating
food waste alone from the amount of biogas produced
from reactors treating food waste and plastic product.
The results are shown in Fig. 3. The calculated biogas
yields were 0.836, 0.187, 0.104, 0.061, 0.053, 0.024
and 0.00 L/gVS of film 2, film 1, film 4, plastic, film 3,
straw, and cup, respectively. An average biogas yield
of 0.73 L/gVS was obtained from food waste digestion
after 43 days. The high degradation of film 2 (PHAs)
may be attributed to high microbial activity and the
contact between microorganisms and the polymers [9].
While the lower biogas yields of other biodegradable
plastics may be attributed to the composition of these
products and the presence of additives used during their
manufacturing. These additives might hinder the
biodegradation [18]. The average final pH of the
reactors were 6.33, 6.77, 6.73, 6.72, 6.74, 6.77, 6.82,
6.87, respectively, for film 2, film 1, film 4, plastic,
film 3, straw, cup, and food waste.
Methane contents of the biogas produced are shown
in Fig. 4. For all reactors containing the food waste and
testing products, except for film 2, methane content
increased over the digestion time from about 25% in
the first week to about 65%-70% at the end of the
second week. After that, the methane content was
almost constant until the end of the digestion time. For
the reactor containing film 2, the methane content
increased, also reaching the same level of 65% at the
end of the second week. It then slightly decreased to
about 60%. This reduction in methane content may
indicate that the digestion of film 2 started after the
depletion of food waste. This may also be confirmed
from the data shown in Fig. 1. The cumulative biogas
for all biodegradable plastics was almost the same
during the first two weeks of digestion and then the
Fig. 4 Methane contents of biogas produced from all
plastic products and food waste: Y error bars are the
standard deviations.
 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products 113

differences in the biogas yields of different test content of the biogas produced. This may be due to the
products became evident. lower F/M ratio at lower initial loading of food waste.

3.3 Biogas Production from Plates and Kraft Paper

The cumulative biogas production yields from the

reactors treating food waste and plates and Kraft paper
are shown in Fig. 5 together with biogas yield from
food waste alone. The reactors treating the plates and
Kraft paper started with the same amount of sludge and
2 gVS/L of plates or Kraft paper together with 1 gVS/L
of food waste. Similar to the reactors treating the
biodegradable plastics products, biogas production
started immediately after the experiments commenced.
The biogas production increased until day 15 for food Fig. 5 Biogas yields of food waste and Kraft paper and
plates.
waste and Kraft paper, and then leveled off. While for
the reactors containing the plates, biogas production
was still increasing till about day 33. Biogas production
rates are shown in Fig. 6. There was a decreased biogas
production rate from day 3 to day 6. Then there was a
peak biogas production rate at the end of the first week
for both Kraft paper and plates. Afterward biogas
production rates decreased sharply. Control reactors
containing 1 gVS/L of food waste were also incubated
to subtract the biogas produced from food waste and
inoculum. The biogas yields from Kraft paper and
plates after excluding the biogas yields from food
waste were 0.133 and 0.507 L/gVS, respectively, as Fig. 6 Biogas production rates from food waste and Kraft
paper and plates.
shown in Fig. 3. The average final pH was 6.73, 6.84
and 6.98 for the reactors containing plates and Kraft
paper together with food waste, and food waste alone,
respectively.
The methane contents of the biogas produced from
reactors treating food waste and either plates or Kraft
paper are shown in Fig. 7. For all the reactors, methane
contents increased over the digestion time till about day
12 and then leveled off. The methane content of the
biogas produced from the reactors containing plates
and paper was lower than that obtained from food
waste reactors. The data in Figs. 4 and 7, for the Fig. 7 Methane content of biogas produced from Kraft
digestion of food waste alone, show that the lower paper, plates and food waste. Y error bars are the standard
initial loading of food waste, the higher methane deviations.
114 Anaerobic Biodegradability of Selected Biodegradable Plastics and Biobased Products
4. Conclusions
The anaerobic biodegradability of six commercial
biodegradable plastics and two cellulose-based
products was evaluated using thermophilic batch
digestion experiments. These products were digested
together with food waste, which served as a source for
the nutrients needed for microbial growth. The results
showed that the degradation of biodegradable plastics
started after the depletion of the biodegradable fraction
of food waste. Only the plastic film made of
polyhydroxyalkanoates and the plate made from sugar
cane bagasse had good digestibility, with biogas yields
of 0.84 and 0.51 L/gVS, respectively, after 43 days of
digestion. Other tested products showed lower or little
biogas yields. After a digestion time of 20 days, little
biodegradation of the studied products was evidence.
This means that in a thermophilic digester treating
municipal solid waste, even at 20 days hydraulic
retentions time, little degradation of biodegradable
plastics will be obtained. At least 43 days are required
to digest most of polyhydroxyalkanoates. This leads to
the conclusion that other technologies should be used
for disposal of non-recycled biodegradable plastic.
Acknowledgments
The authors would like to thank the California
Integrated Waste Management Board for providing the
funding for the project. This research is the
contribution No. 2011-012 of the Cooperative
Research Programs, Lincoln University in Missouri,
Jefferson City, Missouri 65101-0029.
References
[1] M. Shimao, Biodegradation of plastics, Current Opinion in
Biotechnology 12 (2001) 242-247.
[2] Franklin Associates Ltd., Characterization of municipal
solid waste in the United States, 1960 to 2000 (Update
1997), U.S. Environmental Protection Agency,
http://www.epa.gov/garbage/pubs/msw97rpt.pdf.
[3] X. Ren, Biodegradable plastics: A solution or a challenge,
Journal of Cleaner Production 11 (2003) 27-40.
[4] E.S. Stevens, Green Plastics: An Introduction to the New
Science of Biodegradable Plastics, Princeton University
Press, Princeton, New Jersey, USA, 2002.
[5] R. Mohee, G.D. Unmar, A. Mudhoo, P. Khadoo,
Biodegradability of biodegradable/degradable plastic
materials under aerobic and anaerobic conditions, Waste
Management 28 (2008) 1624-1629.
[6] C. Rivard, L. Moens, K. Roberts, J. Brigham, S. Kelley,
Starch esters as biodegradable plastics: Effect of ester
group chain length and degree of substitution on anaerobic
biodegradation, Enzyme and Microbial Technology 17
(2005) 848-852.
[7] M. van der Zee, Biodegradability of
polymers-mechanisms and evaluation methods, in: C.
Bastioli (Ed.), Handbook of Biodegradable Polymers,
Smithers Rapra Technology LTD, UK, 2005.
[8] T. Ishigaki, W. Sugano, A. Nakanishi, M. Tateda, M. Ike,
M. Fujita, The degradability of biodegradable plastics in
aerobic and anaerobic waste landfill model reactors,
Chemosphere 54 (2004) 225-233.
[9] S. Lee, J. Yu, Production of biodegradable thermoplastics
from municipal sludge by a two-stage bioprocess, Resources,
Conservation and Recycling 19 (1997) 151-164.
[10] V. Massardier-Nageotte, C. Pestre, T. Cruard-Pradet, R.
Bayard, Aerobic and anaerobic biodegradability of
polymer films and physico-chemical characterization,
Polymer Degradation and Stability 91 (2006) 620-627.
[11] H. Yagi, F. Ninomiya, M. Funabashi, M. Kunioka,
Anaerobic biodegradation tests of poly (lactic acid) and
polycaprolactone using new evaluation system for
methane fermentation in anaerobic sludge, Polymer
Degradation and Stability and Stability 94 (2009)
1397-1404.
[12] J.B. Van Lier, Thermophilic anaerobic waste water
treatment, temperature aspects and process stability, Ph.D.
Thesis, Wageningen University, the Netherlands, 1995.
[13] R. Zhang, H.M. El-Mashad, K. Hartman, F. Wang, G. Liu,
Characterization of food waste as feedstock for anaerobic
digestion, Bioresource Technology 98 (2007) 929-935.
[14] Standard Methods for the Examination of Water and
Wastewater, 18th ed., APHA (American Public Health
Association), Washington DC, USA, 1998.
[15] ExcelPlas Australia, The impacts of degradable plastic
bags in Australia [Online], 2003,
http://www.deh.gov.au/settlements/publications/waste/de
gradables/impact/pubs/degradables.pdf.
[16] Standard Test Method for Ash Content in Plastics, ASTM
(American Society for Testing and Materials), 2001.
[17] R. Zhang, H.M. El-Mashad, Biogas production from
co-digestion of dairy manure and food waste, Bioresource
Technology 101 (2010) 4021-4028.
[18] K. Sudesh, Y. Doi, Polyhydroxyalkanoates, in: C. Bastioli
(Ed.), Handbook of Biodegradable Polymers, Smithers
Rapra Technology LTD, UK, 2005.
Journal of Environmental Science and Engineering A 1 (2012) 115-120
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Comparison of Antioxidant Activity in Extracts of Myrtus

communis L. Obtained by SFE vs. Solvent Extraction

Paula Pereira1, 2, Maria-João Cebola1, 2 and M. Gabriela Bernardo-Gil1

1. Institute for Biotechnology and Bioengineering, Technical University of Lisbon, Lisboa 1049-001, Portugal
2. Faculty of Engineering and Natural Sciences, Lusophone University of Humanities and Technology, Lisboa 1749-024, Portugal

Received: May 23, 2011 / Accepted: August 24, 2011 / Published: January 20, 2012.

Abstract: This paper reports results of a study on a shrub typical of the Portuguese flora, Myrtus communis L., commonly known as
myrtle. This plant is described in the literature as an antioxidant (AO) source for nutraceutical purposes. We are interested in
assessing its antioxidant capacity (AOC) and to that end we are using the supercritical fluid extraction (SFE) technique to obtain the
plant extracts to study the performance of this technique against the more conventional extraction techniques, namely
hydrodistillation followed by liquid-liquid extraction (LLE) and solid-liquid extraction (SLE). Supercritical extraction studies were
performed using carbon dioxide as the supercritical fluid at different experimental conditions. Cumulative extraction curves for
different temperatures, 308 and 321, 333 K, and pressures 10, 15 and 30 MPa and at constant carbon dioxide flow of about 7 × 10-5
kg·s-1 are presented. The AO activity was determined by three different methods: Folin-Ciocalteu, TEAC (Trolox equivalent
antioxidant capacity) and ORAC (oxygen radical absorbance capacity). The results for the extracts obtained by SFE were compared
with those obtained by conventional extraction showing that SFE is effective to extract AO components from Myrtus communis L.,
and the use of ethanol as co-solvent, at the studied conditions, increases the values of TEAC and ORAC of about four times, when
compared to values obtained when conventional extraction is used.

Key words: Myrtus communis, supercritical fluid extraction, antioxidant activity, Folin-Ciocalteu, TEAC, ORAC.

1. Introduction -tocopherol have been isolated from myrtle extract.

Antioxidants are compounds that react with free
Myrtle (Myrtus communis L.) is an evergreen shrub,
radicals, neutralizing them and thereby preventing or
belonging to the family of Myrtaceae, typical of the
reducing their damaging effects in the body. They
Mediterranean flora. In Portugal, myrtle grows wild
play a major role in our daily life as it is well known
mainly in the central and southern parts of the country.
that oxidative stress is involved in the pathogenesis of
Its leaves are very fragrant, which is the reason for the
numerous diseases. Lipid oxidation is also responsible
extensive use of the plant in the perfume and cosmetic
for deterioration of fats and oils resulting in colour,
industries. It is traditionally used as an antiseptic,
flavour and nutritive value changes. In order to retard
disinfectant and hypoglycemic agent [1]. Different
oxidation, one of the main practices is the addition of
parts of the plant find various uses in the food industry,
antioxidants to fats, fatty foods and cosmetics. The
such as flavoring meat and sauces [2]. Many plants
antioxidants can be of synthetic or natural origin. The
from Myrtaceae family are reported to have
use of synthetic antioxidants is restricted in several
antibacterial or antifungal activities [3] and various
countries, because of their undesirable long-term
known antioxidants like flavonoids, tannins and
toxicological effects, including carcinogenicity [4, 5].
As a result, there is great interest in finding
Corresponding author: M. Gabriela Bernardo-Gil,
associate professor, Ph.D., main research fields: separation antioxidants from natural sources for food and
processes, supercritical fluid extraction, essential oils, medicinal applications.
antioxidants, modeling. E-mail: gbernardo.gil@ist.utl.pt.
116 Comparison of Antioxidant Activity in Extracts of Myrtus communis L. Obtained
by SFE vs. Solvent Extraction
Extracts obtained from herbs and spices have been
described to present antioxidant (AO) activity, and
their use, instead of synthetic antioxidants, is
becoming increasingly popular. Extensive research
has been dedicated to identification of antioxidant
(AO) compounds of natural sources, and the AOC of
many plants has been investigated [5-18]. However,
only rosemary and sage are commercially available as
flavorless, odorless, and colorless AO extracts. Several
studies show that the majority of AO active substances
occurs in extracts, but some essential oils were also
studied to analyze their AO activity, namely essential
oil containing thymol and carvacrol and other phenolic
compounds which exhibit high AO activity.
Romani et al. [19, 20] identified polyphenols in leaf
extracts of Myrtus communis L. Demo et al. [21] used
hexane to obtain extracts of several plants including
myrtle, and they found that their AO activity could be
correlated to the α-tocopherol content. Appendino et
al. [22] obtained extracts from myrtle by using
acetone, having verified that the extracts presented
strong AO activity. In this study, the solvent ability of
a supercritical fluid to extract AO components from a
vegetal matrix is being tested. In SFE the operating
conditions, temperature, pressure and CO2 flow
influence the composition of the extract obtained.
Therefore, a previous optimization study is always
necessary to identify the best operating conditions.
Part of that work is reported here, with studies carried
out at two different temperatures and three different
pressures. The use of a co-solvent was also tested. The
results are shown against those obtained by Clevenger
distillation followed by LLE and SLE applied to the
liquid aqueous and solid extracts obtained from it.
There are several methods for the evaluation of AO
action on fats and oils [23, 24]. In this work, three
different methods of assessing the AOC were used:
the Folin-Ciocalteu method, the TEAC (Trolox
equivalent antioxidant capacity) assay and the ORAC
(oxygen radical absorbance capacity) assay. They
were chosen among the many methods available
because overall they give a fairly accurate account of
the AOC of a plant extract [25].
The Folin-Ciocalteu method (F-C) measures the
total content in polyphenols and assumes that it is
proportional to the AO activity. The results are
calculated as gallic acid equivalents [26].
The TEAC assay is based on the generation of the
ABTS (2,2’-azinobis(3-ethylbenzothiazoline-6-
sulfonic acid)) radical cation (ABTS+) and by the
measurement of its decay caused by the addition of an
AO containing sample. The results are reported as
trolox equivalents [27].
The ORAC assay measures the free radical
oxidation of a fluorescent probe by the change in its
fluorescence intensity over time, following the
reaction (through the loss of fluorescence) for
extended periods of time (60 min or more).
Calculation of AO activity is based on the net
integrated areas under the fluorescence decay curves.
The results are calculated as trolox equivalents [28].
2. Materials and Methods
2.1 Plant Material
Samples of myrtle leaves were collected from the
Sintra area (Portugal). The plant material, after
collection, was identified and deposited in the
Herbarium of the Superior Agronomy Institute of the
Technical University of Lisbon (code: LISI). The
plant material kept for analysis was dried for two
months out of the sunlight, and then sealed in black
bags and kept at -4 ºC.
2.2 Chemicals
Diisopropylether, used in liquid-liquid extraction,
was of p.a. quality and purchased from Merck kGaA
(Darmstadt, Germany). Folin-Ciocalteu reagent, and
gallic acid were from Fluka (Buchs, Switzerland),
sodium carbonate from VWR (Leuven, Belgium),
trolox standards (0.1-0.4 mM) from Calbiochem
(Darmstadt, Germany), potassium persulfate and
ABTS (2,2’-azinobis (3-ethylbenzothiazoline-
 Comparison of Antioxidant Activity in Extracts of Myrtus communis L. Obtained
by SFE vs. Solvent Extraction
6-sulfonic acid)) from Sigma-Aldrich (Steinheim,
Germany), fluorescein from Panreac (Spain,
Barcelona), and AAPH (2,2'-azobis-
2-methyl-propanimidamide, dihydrochloride) from
Acros (Geel, Belgiím).
The carbon dioxide used was CO2 N45 (purity 
99.995%) and supplied by AIR LIQUIDE (Lisboa,
Portugal).
2.3 Hydrodistillation
The essential oil (EO) was obtained by Clevenger
distillation and analyzed by GC and GC/MS. The
results show that the major components are
limonene+1,8-cineole, myrtenyl acetate, -pinene and
linalool [29].
The aqueous phase obtained from the Clevenger
distillation and the remaining solid residues were
subjected to two different extraction procedures: the
aqueous phase was extracted with diisopropylether
obtaining what we designate as liquid phase extract
(LLE). The solid residue was first extracted with
boiling water and the liquid phase thus obtained was
subsequently extracted with diisopropylether obtaining
what we designate as solid phase extract (SLE). Both
extracts were subject to the determination of AOC
using the three chosen methods: the Folin-Ciocalteu,
the TEAC assay and the ORAC assay.
2.4 Supercritical Fluid Extraction
Supercritical extraction studies were performed in a
semi-batch flow extraction apparatus similar to that
referred elsewhere [7, 8], by using carbon dioxide as
the supercritical fluid. The extraction cell, with a
volume of 278 cm3, was placed inside a
temperature-controlled oven. The extract was obtained
in two stages: first, the heavier components were
collected at ambient temperature, and the lighter
components were obtained in a trap under cryogenic
conditions and at atmospheric pressure. Ethanol was
used as co-solvent. The experimental conditions that
were used in the supercritical fluid extraction are
117
presented in Table 1.
The solid residues from the supercritical extractions
were also subjected to solid liquid extraction with water
and the aqueous phase extracted subsequently with
diisopropylether, achieving extracts named SFE Res.
All the SFE extracts and the SFE Res extracts were
also subjected to the determination of AO activity, by
using the three chosen methods: the Folin-Ciocalteu,
the TEAC and the ORAC assays.
2.5 AOC Assays
2.5.1 Folin-Ciocalteu (FC) Assay
The total content in polyphenols was obtained by
the F-C colorimetric method reported by Ainsworth et
al. [26], with the modifications described by
Grevenstuk et al. [30] were used. The absorbance was
measured at 765 nm in by using an Infinite 200 (Tekan,
Grodig, Austria) microplate reader. The results were
expressed as mmol gallic acid per gram plant.
2.5.2 TEAC (Trolox Equivalent Antioxidant
Capacity) Assay
The antioxidant activity was evaluated by the
TEAC assay using the radical cation ABTS+,
according the procedure of Re et al. [27], and
described by Grevenstuk et al. [30]. The absorbance
was measured at 734 nm by using an Infinite 200
microplate reader (Tekan, Grodig, Austria). The
results were expressed as mmol trolox per gram plant.
2.5.3 ORAC (Oxygen Radical Absorbance Capacity)
Assay
The total antioxidant capacity was analyzed by the
ORAC method described by Gillespie et al. [28]. The
decrease in fluorescence of fluorescein was evaluated
by reading for exciting at 485 nm, and emission at 530
nm. ORAC results were expressed as mmol trolox per
gram plant.
3. Results and Discussion
Cumulative extraction curves for the different
conditions can be seen in Figs. 1 and 2. Fig. 2 was
introduced for a better assessment of the yields
118 Comparison of Antioxidant Activity in Extracts of Myrtus communis L. Obtained
by SFE vs. Solvent Extraction

Table 1 Experimental conditions of SFE experiments.

Run P (MPa) T (K) CO2 density (kg·m-3) Co-solvent ethanol (%)
SFE1 10 333 290.00 -
SFE2 30 308 929.11 -
SFE3 15 321 717.14 -
SFE4 10 321 421.63 5

14 far higher than all the other yields. This is not

12
S FE1
surprising given that carbon dioxide is a non-polar
S FE2
10 S FE3
solvent and the components that we aim to extract are
Yield (%)

8 S FE4 polar in its nature.

6 The results of TEAC and ORAC methods,
4 expressed in term of mmol trolox/g plant, obtained for
2 all the SFE extracts are presented in Fig. 3, as well as
0 the values of total phenolics obtained by the
0 200 400
Time (min)
600 800 Folin-Ciocalteu method, expressed as mmol gallic
Fig. 1 Comparison between extract yields obtained for acid/g plant. In the same figure, the values obtained
different SFE conditions. for the SFE solid residues and those obtained by
conventional extraction are also presented.
2.0
The results obtained for the AO activity follow the
1.5 same trend as for yield obtained in SFE. In the SFE1
S FE1
Yield (%)

S FE2
1.0
S FE3
0.5
0.0
0 200 400 600
Time (min)
Fig. 2 Comparison between extract yields obtained for
different SFE conditions excluding the experiment where
co-solvent was used.
obtained in the SFE experiments where the co-solvent
was not used, as they are much lower than the yield
obtained when the co-solvent was used.
The number of SFE experimental runs so far is still
insufficient to reach definite conclusions as far as the
pressure and temperature parameters are concerned.
Nevertheless, the results plotted in Fig. 2 seem to
show that better yields are achieved at higher
pressures and low temperatures, i.e, the yield rises as
the density increases, the opposite resulting in very
low recoveries. But the marked effect of the use of a
co-solvent is obvious (Fig. 1): the extraction yield is
experiment the AOC was very low for all the methods
and the results of the AOC in the solid residue show
higher activity. This proves that the extraction of the
AO components was not totally achieved as they
remained in the solid residue. In SFE3 the AOC
800
increased in the extract compared to SFE1 but it is
still the solid residue that shows the highest activity.
In agreement with the yield cumulative curves, the
SFE2 results show better AOC in the extract. Overall,
the results obtained by each of the three methods used
are in agreement, but the TEAC values exhibit a closer
trend with F-C then ORAC does. In the case of SFE2,
ORAC values are higher for the extract than for the
solid residue while the opposite occurs for TEAC and
F-C. This reveals the different mechanisms for AOC
determination that are used for the different assays.
TEAC and ORAC results obtained for SFE 4 extract
are much higher than the one for SFE 4 solid residue,
showing that ethanol is a good co-solvent for
increasing the extraction of AO components. All the
extracts obtained by SFE have higher ORAC values
 Comparison of Antioxidant Activity in Extracts of Myrtus communis L. Obtained 119
by SFE vs. Solvent Extraction

250 100
TEAC
90

mmol gallic acid/g plant (F-C)

ORAC
200 F-C 80
mmol Trolox/ g plant

70

150 60

50

100 40

30

50 20

10

0 0
SFE1 Ext SFE1 Res SFE2 Ext SFE2 Res SFE3 Ext SFE3 Res SFE4 Ext SFE4 Res Clev LLE ClevE SLE

Fig. 3 Comparison between the antioxidant activity of the SFE and Clevenger residues extracts measured by TEAC, ORAC,
and Folin-Ciocalteu (F-C).

than TEAC and F-C, while the reverse is true for the 613-617.
[3] O. Gortzi, S. Lalas, I. Chinou, J. Tsaknis, Reevaluation of
extracts obtained by solvent extraction after Clevenger
bioactivity and antioxidant activity of Myrtus communis
distillation. extract before and after encapsulation in liposomes, Eur.
Food Res. Technol. 226 (2008) 583-590.
4. Conclusion [4] M. Namiki, Antioxidant/antimutagens in food, CRC Crit
Rev. Sci. Nut. 29 (1990) 273-300.
SFE was applied to ground myrtle leaves and the
[5] G. Gazzani, A. Papetti, G. Massolini, M. Daglia, Anti-
extracts were studied in terms of antioxidant activity. and prooxidant activity of water soluble components of
Folin-Ciocalteu, TEAC, and ORAC follow the same some common diet vegetables and the effect of thermal
tendency as the yield has. Although the work is treatment, J. Agric. Food Chemistry 46 (1998)
4118-4122.
continuing, it is evident from these experiments that [6] Z.Y. Chen, L.Y. Wang, P.T. Chan, Z. Zhang, H.Y. Chung,
the yield and the AO activity are higher when ethanol C. Liang, Antioxidative activity of green tea catechin
is used as co-solvent. extract compared with that of rosemary extract, JAOCS
75 (1998) 1141-1145.
Acknowledgments [7] M.M. Esquível, M.A. Ribeiro, M.G. Bernardo-Gil,
Supercritical extraction of savory oil: Study of
Authors thank Dr. T. Grevenstuk, P. Costa, Dr. S. antioxidant activity and extract characterization, J.
Gonçalves and Prof. A. Romano of Laboratory of Supercritical Fluids 14 (1999) 129-138.
[8] M.A. Ribeiro, MG. Bernardo-Gil, M.M. Esquível,
Plant Biotechnology of University of Algarve for help
Melissa Officinalis L.: Study of antioxidant activity in
in AOC analysis. supercritical residues, J. Supercritical Fluids 21 (2001)
51-60.
References [9] V. Reddy, A. Urooj, A. Kumar, Evaluation of antioxidant
[1] M.S. Elfellah, M.H. Akhter, M.T. Khan, activity of some plant extracts and their application in
Antihyperglycaemic effect of an extract of Myrtus biscuits, Food Chemistry 90 (2005) 317-321.
communis in Streptozotocin-induced diabetes in mice, J. [10] N.V. Yanishlieva, E. Marinova, J. Pokorný, Natural
Ethnopharmacology 11 (1984) 275-281. antioxidants from herbs and spices, Eur J. Lipid Sci.
[2] J. Chalchat, R. Garry, A. Michet, Essential oils of myrtle Technol. 108 (2006) 776-793.
of the Mediterrenean litoral, J. Essent. Oil Res. 10 (1998) [11] M. Rubilar, M. Pinelo, M. Ihl, E. Scheuermann, J. Sineiro,
120 Comparison of Antioxidant Activity in Extracts of Myrtus communis L. Obtained
by SFE vs. Solvent Extraction
M.J. Nuñez, Murta leaves (ugni molinae turcz) as a
source of antioxidant polyphenols, J. Agric. Food Chem.
54 (2006) 59-64.
[12] N. Erdemoglu, N.N. Turan, I. Cakici, B. Sener, A. Aydm,
Antioxidant activities of some lamiaceae plant extracts,
Phytot. Res. 20 (2006) 9-13.
[13] M. Natsume, Y. Muto, K. Fukuda, T. Tokunaga, N.
Osakabe, Determination of rosmarinic acid and luteolin
in Perilla frutescens Britton (Labiatae), J. Sci. Food
Agric. 86 (2006) 897-901.
[14] M. Pérez-Mateos, C. Lanier, L. Boyd, Effects of
rosemary and green tea extracts on frozen surimi gels
fortified with omega-3 fatty acids, J. Sci. Food Agric. 86
(2006) 558-567.
[15] I. Amin, O. Mukhrizah, Antioxidant capacity of
methanolic and water extracts prepared from
food-processing by-products, J. Sci. Food Agric. 86
(2006) 778-784.
[16] R.Y. Yang, S.C.S. Tsou, T.C. Lee, W.J. Wu, P.M.
Hanson, G. Kuo, et al., Distribution of 127 edible plants
species for antioxidant activities by two assays, J. Sci.
Food Agric. 86 (2006) 2395-2403.
[17] L. Martínez, I. Cilla, J.A. Beltrán, P. Roncalés,
Antioxidant effect of rosemary, borage, green tea, pu-erh
tea and ascorbic acid on fresh pork sausages packaged in
a modified atmosphere: Influence of the presence of
sodium chloride, J. Sci. Food. Agric. 86 (2006)
1298-1307.
[18] N. Pellegrini, B. Colombi, S. Salvatore, O.V. Brenna, G.
Galaverna, D. Del Rio, et al., Evaluation of antioxidant
capacity of some fruit and vegetable foods: Efficiency of
extraction of a sequence of solvents. J. Sci. Food Agric.
87 (2007) 103-111.
[19] A. Romani, P. Pineli, N. Mulinacci, F.F. Vincieri, M.
Tattini, Identification and quantification of polyphenols
in leaves of Myrtus communis L., Chomatographia 49
(1999) 17-20.
[20] A. Romani, R. Coinu, S. Carta, P. Pinelli, C. Galardi, F.
Vincieri, et al., Evaluation of antioxidant effect of
different extracts of Myrtus communis L., Free Radical
Res. 38 (2004) 97-103.
[21] A. Demo, C. Petrakis, P. Kefalas, D. Boskin, Nutrient
antioxidants in some herbs and Mediterranean plant
leaves, Food Res. Int. 31 (1999) 351-354.
[22] G. Appendino, F. Bianchi, A. Minassi, O. Sterner, M.
Ballero, S. Gibbons, Oligomeric acylphloro glucinols
from myrtle (Myrtus communis), J. Nat. Prod. 65 (2002)
334-338.
[23] E.N. Frankel, A.S. Meyer, The problems of using
one-dimensional methods to evaluate multifunctional
food and biological antioxidants, J. Sci. Food Agric. 80
(2000) 1925-1941.
[24] E.M. Becker, L.R. Nissen, L.H. Skibsted, Antioxidant
evaluation protocols: Food quality or health effects, Eur.
Food Res. Technol. 219 (2004) 561-571.
[25] R.L. Prior, X. Wu, K. Schaich, Standardized methods for
the determination of antioxidant capacity and phenolics in
food and dietary supplements, J. Agric. Food Chem. 53
(2005) 4290-4302.
[26] E.E. Ainsworth, K.M. Gillespie, Estimation of total
phenolic content and other oxidation substrates in plant
tissues using Folin-Ciocalteu reagent, Nature Protocols 2
(2007) 875-877.
[27] R. Re, N. Pellegrini, A. Proteggente, A. Pannala, M.
Yang, C. Evans, Antioxidant activity applying an
improved ABTS radical cation decolorization assay, Free
Radical Biology Medicine 26 (1999) 1231-1237.
[28] K.M. Gillespie, J.M. Chae, E.A. Ainsworth, Rapid
measurement of total antioxidant capacity in plants,
Nature Protocols 2 (2007) 867-870.
[29] P.C. Pereira, M.J. Cebola., M.G. Bernardo-Gil, Evolution
of the essential oil composition of Portuguese myrtle
through the vegetative cycle, Molecules 14 (2009) 1-14.
[30] T. Grevenstuk, S. Gonçalves, S. Almeida, N. Coelho, C.
Quintas, M.N. Gaspar, et al., Evaluation of antioxidant
and microbial properties of in vitro cultured drosera
intermedia extracts, Natural Product Communications 4
(2009) 1-6.
Journal of Environmental Science and Engineering A 1 (2012) 121-127
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Feasibility of Using Sea Shells Ash as Admixtures for

Concrete

Benjamin R. Etuk1, Idongesit F. Etuk1 and Linus O. Asuquo2

1. Department of Chemical and Petroleum Engineering, University of Uyo, Uyo 52001, Nigeria
2. Department of Mechanical Engineering, University of Uyo, Uyo 52001, Nigeria

Received: June 24, 2011 / Accepted: August 2, 2011 / Published: January 20, 2012.

Abstract: This research aims at producing pozzolanic admixtures from waste shells of periwinkle, oyster and snail. The clean shells
were ashed at 800 °C and analysed for chemical composition. The results show that they are pozzolanic in nature. The effect of using
the shells ash as admixtures on the setting time and compressive strength of cement paste and mortar were investigated using varying
percentages of 0%, 5%, 10%, 15%, 20%, 25% and 30% by weight of each of the shell ashes. The results show that the water
consistency, initial, and final setting times of each of the cement blends were increasing with increase in percentage replacement of
cement with each of the shell ashes. The optimum compressive strength with percentage replacement level of 10% for periwinkle
shell ash (PSA), 15% for oyster shell ash (OSA), and 20% for snail shell ash (SSA) were obtained with ashes produced.

Key words: Sea shell ash, concrete, admixtures, waste utilization.

1. Introduction available and/or waste materials to partially or fully

replace the costly conventional materials.
Housing delivery in Nigeria in the rural, semi-urban
An admixture is a material other than cement, water
and urban areas is still a serious problem due partly to
and aggregates that is used as an ingredient of
the high cost of some essential materials, especially
concrete and is added to the batch immediately before
cement (a major material in the construction industry)
or during mixing [2]. According to Michael [3],
whose cost has increased over the years and may soon
admixtures in concrete can improve its workability,
be out of reach of the ordinary citizens. Recently, to
hardening, or strength characteristics and generally
arrest the escalating cost of cement, the Federal
result in a reduction in the cost of concrete
government directed local manufacturers to evolve
construction [4]. For instance, activated kaolin clay,
strategies to bring down the price of the commodity to
fly ash, silica fume, baryte powder and oil shale ash
avoid a review of the existing policy on cement which
had been used as components of blended binders for
gave 2013 as the deadline for the country to be
mortars with good results in terms of their pozzalinic
self-sufficient in cement production [1].
activity [5]. The use of cassava starch [6], rice husk
Concrete is an essential construction material
ash [7-10], rice straw ash [11], and wood waste ash
composed of cement, aggregate (gravel or granite and
[12] to partially replace different types of cement in
sand) and water. In order to reduce the construction,
concrete had also been reported. The use of these
researchers have been exploring the possibility of
agricultural materials helped to reduce waste as well
using pozzolanic admixtures produced from locally
as improve the strength properties of the concrete.
However, the strength reduced with time due to
Corresponding author: Benjamin R. Etuk, associate
professor, Ph.D., main research fields: separation processes, microbial activity.
particulate systems, food processing, waste control and In Nigeria, the intensity of fishing activities in the
utilization. E-mail: bretuk2001@yahoo.com.
122 Feasibility of Using Sea Shells Ash as Admixtures for Concrete
riverine areas has resulted in the production and
accumulation of large quantities of sea shells as
wastes along the coastal regions, market areas and
dump sites. Usually, the fleshes of the species are
processed for consumption while the inedible hard
shells are dumped at open sites thereby causing
environmental pollution.
Periwinkle, snail, oyster and all marine animals
belonging to phylum mollusca and class gastropoda
[13]. They belong to the group of exoskeleton animals.
The exoskeletons contain rigid and resistant
components that fulfill a set of functional roles
including protection, excretion, support, feeding,
acting, etc.. They contain chitin and when calcium
carbonate is added, the exoskeleton grows in strength
and hardness [14].
An exploratory study on the suitability of sea shells,
such as periwinkle shells, as partial or full
replacement for granite in concrete work had been
reported with satisfactory compressive test results at
appropriate concrete mix ratios [15]. Using the ash of
the sea shell, investigation had shown that up to 50%
replacement of cement in sandcrete blocks and 5%
replacement in laterite blocks were possible with good
results in terms of compressive strength [16, 17].
In this paper, the utilization of the ash of waste
shells of periwinkle (PSA), snail (SSA) and oyster
(OSA) as partial replacement for cement in the
construction industry is reported. The concentrations
of the ash from the different shells in finely divided
form are evaluated to determine the optimum needed
that will not compromise the workability, hardening
and strength properties of the concrete produced. The
effective utilization of these sea shell wastes which are
available almost free of cost and in abundance will not
only reduce their pollution tendency but will help in
reducing the amount of cement used in concrete work.
2. Materials and Methods
The materials used in carrying out the research were
periwinkle, snail and oyster shells. Other materials
used were UNICEM® Ordinary Portland Cement
(OPCEM), water and sand.
The periwinkle, snail and oyster shells were
obtained from Akpan Andem market in Uyo, a dump
site at Okpoedu, Itu, and Issiet in Itu and Uruan Local
Government Areas respectively, all in Akwa Ibom
State, Nigeria.
2.1 Production of Pozzolanic Admixture
The three different shell samples were each washed
thoroughly to remove dirt and mud and then sun-dried
for three days. The samples were then placed in a
furnace and ashed at temperatures of 800 °C for a
period of four hours. The ash samples were then
ground into a powdered form using metallic mortar
and pestle, sieved through a sieve mesh size of 63
microns, and kept in tight containers for analyses.
2.2 Chemical Analyses
The shell ashes were analysed to determine their
composition at the Quality Control Laboratory of
Ashakacem, Gombe, State, in accordance with
Nigerian Industrial Standard (NIS) method [18]. In the
method, stearic acid (0.4 g) and 20.0 g of each of the
three different shell ashes obtained at temperatures of
800 °C were measured and were put in a grinding pot
and ground for 60 seconds using the Herzog grinding
machine. The aluminum cup was filled half way with
stearic acid and then filled up with the samples. The
cup was carefully inserted into the pellet making
machine (Herzog pressing machine). The pellet
produced was then placed in the cement X-ray
spectrophotometer and the programme to which the
samples were analysed was selected, that is, the
program OPCEM (ordinary Portland cement), and the
start key was clicked so that within 2-3 minutes the
results were obtained [19].
2.3 Determination of Specific Gravity
The specific gravity of the ash samples was
determined in accordance with BS method [20]. The
 Feasibility of Using Sea Shells Ash as Admixtures for Concrete
empty density bottle with stopper was weighed as W1
and then filled with shell ash to about three quarter of
the bottle. This was measured as W2. The bottle
containing the shell ash was then filled up with water
and a stopper used to cover it. This was measured as
W3. The content of the density bottle were then poured
out and the bottle rinsed with water. The bottle was
thereafter filled with water and the stopper inserted.
This was measured as W4. The specific gravity for
each of the shell ashes was determined, using the
formula:
W2 – W1
Sp gr = (1)
(W2 – W1) – (W3 – W4)
2.4 Determination of Consistency of Cement Pastes
Four hundred and ninety grams of the OPCEM was
weighed and placed on a non-absorbent metallic tray.
By trial mixtures, the required water content which
produced the cement paste of desired standard
consistency of between 26 and 33 (expressed as a
percentage by mass of the dry cement) was chosen.
The measured water was then added to the weighed
cement, thoroughly mixed and finely ground together
with a hand-trowel for four to five minutes to form a
neat cement paste. The neat cement paste was then
placed in a special metallic mould and the consistency
of the neat cement paste was then determined by
lowering the plunger which is attached to the Vicat®
apparatus and allowed to make contact with the top
surface of the paste before it was finally released.
Under the action of its weight, the plunger was
allowed to penetrate the paste and the depth of
penetration for a standard and consistent cement
paste was to a point 5 mm to 7 mm from the bottom of
the mould. This test meets the requirement stated in
BS [21].
2.5 Determination of Initial and Final Setting Times
The plunger used for consistency test was replaced
with a round needle with a cross-sectional area of 1
mm2 used as the initial set needle. The process of
123
making the needle penetrate the paste of standard
consistency was repeated at intervals of 5 minutes,
until the paste was stiffed enough for the initial set
needle to penetrate only to a point 10 mm to 20 mm
from the bottom of the mould. The initial setting time
was recorded as the time that elapsed from when the
paste was made to when it set [21].
The initial set needle was then replaced with the
final set needle with a 1 mm square needle having a
circular cutting edge of 5mm in diameter. The final set
needle was made to penetrate the paste in the mould,
so that it left a circular cutting edge of 5 mm in
diameter and set 0.5 mm behind the tip of the needle.
The process of allowing the needle penetrate the paste
was repeated at intervals of 5 minutes, and the final
set was said to have taken place when the needle
which was gently lowered to the surface of the paste
made an impression on it but the circular cutting edge
failed to penetrate [6].
2.6 Production of the Mortar Cubes
The moulds of size 50 × 50 × 50 mm were used for
all the casting of the mortar cubes. The moulds were
cleaned and oiled to enhance easy removal of the
cubes after setting and prevent damage of the test
cubes. A mixture of 0.27 kg of cement, 0.81 kg of
sand, and 0.135 kg of water, all in a mix ratio of
(1:3:0.5) was measured. The mixture was mixed
thoroughly by means of a trowel on a non-absorbent
metallic tray to obtain a homogenous mixture. This
was used as a reference sample (i.e 0% replacement).
The mix was then transferred into the mould of 50 ×
50 × 50 mm and filled in three equal layers. Each of
the layers was compacted 25 times using a rod of three
quarter diameter and allowed for twenty four (24)
hours before removing it and then cured in water for
7 days.
This process was repeated with cement replaced by
each of the shell ashes by weight at varied
concentrations of 5, 10, 15, 20, 25 and 30 percent.
The cubes were put into a curing tank containing
124 Feasibility of Using Sea Shells Ash as Admixtures for Concrete

water. This was done to maintain satisfactory moisture cement used. Nevertheless, they are in accordance
content as that hydration of the cementious material with BS [20].
continues long enough to achieve the required strength,
3.3 Consistency/Setting Time Test
durability and reduce shrinkage induced cracking in
the cube [22]. Fig. 1 shows the results of water consistency of
cement paste blended with PSA, OSA and SSA
2.7 Compressive Strength Test
produced at temperature of 800 °C. The water
This test was done in accordance with BS EN consistency of the blended cement paste increases
method [23] and used by Oymael [24]. The cubes were with increase in percentage replacement in the
removed from the curing tank at the end of the curing following trend of PSA > OSA > SSA. The reason for
period (7 days) and then weighed, before compressive this trend may be attributed to the high silica contents
strength was conducted. Three cubes each for the in PSA compared to OSA and SSA. For SSA the
different replacement levels were crushed at 7 days lowest water consistency could also be as a result of
using a manual compressive machine, with capacity of high lime content.
1000 kN. The test was carried out in the Building The results of initial and final setting times of
Department Laboratory, University of Uyo, Uyo. cement paste blended with PSA, OSA and SSA
obtained at 800 oC are presented in Figs. 2 and 3. The
3. Results and Discussion
results show that the initial and final setting times
3.1 Chemical Analyses increase with increase in the percentage replacement
of each of the shell ash. This may be due to the
The chemical composition of the OPCEM used and
increase in the required mix water, as well as retarded
the shell ashes are shown in Table 1. The results show
hydration caused by having more of the shell ashes
the cement to have the four major compounds, namely,
than cement in the mix.
CaO, SiO2, Al2O3, and Fe2O3, with high percentages
Table 1 Chemical composition of OPCEM and the sea
of CaO and SiO2 which accounts for its strength. shells ash obtained at 800 °C.
Also the results for each of the shell ashes show that Composition (wt %)
Component
they contain the main chemical compounds of cement OPCEM PSA SSA OSA
namely, CaO, SiO2, Al2O3, and Fe2O3, and the SiO2 20.06 26.26 10.20 13.41
similarity in most of the chemical composition in the Al2O3 5.85 8.79 4.81 4.95
Fe2O3 3.05 4.82 3.15 3.80
OPCEM and the latter make the partial replacement of
CaO 61.44 55.53 61.95 57.95
cement by each of the shell ashes to be feasible. From MgO 0.93 0.4 0.18 0.19
the data presented, the amount of sulphur trioxide SO3 2.71 0.18 0.03 0.12
(SO3) present in each of the shell ashes lies within the K2O 0.97 0.20 0.05 0.02
Na2O 0.14 0.25 0.04 0.22
optimum range of not more than 3.0% recommended
P2O5 0.17 0.05 0.01 0.01
by ASTM [25]. However, the results show that PSA MnO3 0.20 0.07 0.01 0.01
contains more SO3 followed by OSA and SSA. TiO2 0.28 0.05 0.01 0.01

3.2 Specific Gravity Test Table 2 Specific gravity.

Material Specific gravity
Table 2 shows the results of the specific gravity for Cement 3.10
cement, periwinkle shell ash, snail shell ash and oyster Periwinkle shell ash 2.50
shell ash obtained. From the data, it is clear that the Oyster shell ash 2.33
Snail shell ash 2.44
specific gravity of the ashes are all lower than that of
 Feasibility of Using Sea Shells Ash as Admixtures for Concrete 125

Fig. 1 Water consistency with percentage replacement of cement with sea shell ash in concrete.

Fig. 2 Initial setting time with percentage replacement of cement with sea shell ash.

Fig. 3 Final setting time with percentage replacement of cement with sea shell ash.
126 Feasibility of Using Sea Shells Ash as Admixtures for Concrete
Fig. 4 Compressive strength with varying percentage replacement of cement with sea shell ash.
3.4 Compressive Strength
The results of the compressive strength of the
mortar cubes produced from cement blended with
each of the shell ashes are shown in Fig. 4. The figure
shows the plots of the 7 days compressive strength
against percentage replacement of PSA, OSA and
SSA obtained at temperatures of 800 °C. The results
show that at first the compressive strength of each of
the mixes is low compared to that of the control mix.
This is so, because the pozzolanic activity is slow as it
allows for the hydration of cement, but later increases
with increases in percentage replacement up to 10%
by weight of PSA, 15% for OSA and 20% for SSA,
and then decreases as the percentage for each of the
shell ash increases.
4. Conclusions
Based on the results of this study, the following
conclusions are drawn:
(1) Periwinkle shell ash (PSA), oyster shell ash
(OSA) and snail shell ash (SSA) are pozzolanic in
nature and satisfies the requirements of ASTM [25].
Therefore, it can be used as a cement replacement
material;
(2) The water consistency increases with increase in
the percentage replacement of each of the shell ashes;
(3) The initial and final setting times of the blended
cement pastes were found to increase with increasing
percentage replacement of each of the shell ashes;
(4) The compressive strength of the mortar cubes
decreases with increase in the amount of the shell ash
in the cement paste;
(5) The cement can be replaced partially by up to
10% by weight of periwinkle shell ash (PSA), 15% by
weight of oyster shell ash (OSA), and 20% by weight
of snail shell ash (SSA) in making of mortar cubes
without the strength being affected.
Acknowledgments
The authors are deeply thankful to Ashaka Cement
Company, Gombe State for their help in the chemical
analyses of the ash samples.
References
[1] R. Badmus, Cement: Manufacturers fine-tune strategies
to meet ultimatum, The Punch Newspaper (Nigeria), May
20, 2011, p. 32.
[2] M.S. Shetty, Concrete Technology, Theory and Practice,
revised ed., S. Chand and Company Ltd., Ram Nagar,
New Delhi, 2005, pp. 124-217.
[3] R.L. Michael, Civil Engineering Reference Manual, 6th
ed., Professional Publications, Inc., Belmont, CA, USA,
1992, pp. 142-144.
[4] C. Piper, Why Use an Admixture with Concrete?
[Online],
 Feasibility of Using Sea Shells Ash as Admixtures for Concrete
http://www.ehow.com/info_8761265_use-admixture-con
crete.html.
[5] A. Shvarzman, K. Kovler, I. Schamban, G. Grader, G.
Shter, Influence of chemical and phase composition of
mineral admixtures on their pozzolanic activity,
Advances in Cement Research 13 (2001) 1-7.
[6] O.E. Alutu, T.O. Ehondor, The setting time of portland
cement pastes dosed with cassava starch admixtures,
Nigerian Journal of Industrial & Systems Studies 4 (2005)
24-31.
[7] D.C. Okpala, Rice husk ash (rha) as partial replacement
of cement in concrete, in: Proceeding of the Annual
Conference of the Nigerian Society of Engineers, Port
Harcourt, Nigeria, 1987, pp. 20-25.
[8] J.I.C. Mbachu, J.O. Kolawole, Shrinkage and elastic
moduli of ordinary portland cement (opc) and rice husk
ash (rha) concrete made with different coarse aggregate
types, J. Envir. Sci. 1 (1998) 35-40.
[9] K. Sakr, Effects of silica fume and rice husk ash on the
properties of heavy weight concrete, J. Mater. Civ. Eng.
18 (2006) 367-376.
[10] E.B. Oyetola, M. Abdullahi, The use of rice husk ash in
low-cost sandcrete block production, Leornardo Electronic
Journal of Practices and Technologies 5 (2006) 58-70.
[11] M.A. El-Sayed, T.M. El-Samni, Physical and chemical
properties of rice straw ash and its effect on the cement
paste produced from different cement types, J. King Saud.
Univ. Eng. Sci. 19 (2006) 21-30.
[12] F.F. Udoeyo, H. Inyang, D.T. Young, E.E. Paradu,
Potential of wood waste ash as an additive in concrete, J.
Mater. Civ. Eng. 18 (2004) 605-611.
[13] C. Peter, E. Michael, Marine Biology, 15th ed., Mc
Graw-Hill Companies, New York, 2005, pp. 167-225.
[14] B. Nicholas, Exceptional fossil preservation and the
combian explosion, Integrative and Comparative Biology
127
43 (2003) 166-177.
[15] A.P. Adewuyi, T. Adegoke, Exploratory study of
periwinkle shells as coarse aggregates in concrete works,
Journal of Applied Sciences Research 4 (2008)
1678-1681.
[16] A.A. Umoh, I.A. Odesola, The effects of periwinkle shell
filler on the compressive strength developments of
lateritic blocks, Intern. J. Eng. 2 (2008) 1-11.
[17] O.F. Job, A.A. Umoh, S.C. Nsikak, Engineering
properties of sandcrete blocks containing periwinkle shell
ash and ordinary Portland cement, Inter. J. Envir. Sci. 1
(2009) 18-24.
[18] Nigeria Industrial Standard, NIS 412, Standards for
Cement, Standard Organization of Nigeria, Abuja,
Nigeria, 2009, pp. 9-15.
[19] M. Abdulfattah, Determination of chemical analysis
using pellets, in: A laboratory Manual of Quality Control,
Ashakacem Plc., Gombe State, 2001, pp. 1-12.
[20] Methods for Sampling and Testing of Mineral Aggregates,
Sands and Fillers, British Standard Institution, London,
1975.
[21] Specification for Ordinary and Rapid Hardening of
Portland Cement, British Standard Institution, London,
2000.
[22] A.M. Neville, Properties of Concrete, 5th eds., Ritman
Limited, New York, 2000, p. 900.
[23] Cement, Composition, Specifications and Conformity
criteria for Common Cements, British Standards
Institution, London, 1997.
[24] S. Oymael, Suitability of oil shale ash as a constituent of
cement, Oil Shale Journal 24 (2007) 45-58.
[25] Specification for Fly Ash and Raw or Calcined Natural
Pozzolanas for Use as Mineral Admixture in Portland
Cement Concrete, American Society for Testing and
Materials Standards, Philadelphia, USA, 2008.
Journal of Environmental Science and Engineering A 1 (2012) 128-132
Formerly part of Journal of Environmental Science and Engineering, ISSN 1934-8932
D DAVID PUBLISHING

Dazzling Pearl-Analysis of the Artistic Characteristics of

Wu Wei’s TianTi Mountain Grotto

Jing Xue
Department of Design and Art, Lanzhou University of Technology, Lanzhou 730050, China

Received: September 14, 2010 / Accepted: November 8, 2010 / Published: January 20, 2012.

Abstract: With the trend of studying the region culture and non-material culture, people seem to pay no attention on the TianTi
Mountain Grotto. But it plays an important role in the mountain grotto in the central China. This paper analyses the artistic value of
Wu Wei’s TianTi Mountain Grotto through the study of the history, influence and shape characteristics of it.

Key words: TianTi Mountain Grotto, shape characteristic, artistic characteristics.

1. Introduction 2. The History and Influence of the Grottoes

The TianTi Mountain Grotto, and also called
Liangzhou Grotto, is located at 50 kilometers south of
the city. There is a temple at the foot of the cave, which
is called Grand Buddha Temple or Guanshan Temple.
According to Qianlong “Wu Wei County Annal”: “In
the Grand Buddha Temple about a hundred miles
southeast to the city, there is a big buddha statue which
is about thirty meters high”. The temple is also known
as Guanshan Temple [1]. This is also written down in
such Buddhism classics as “Fa Yuan Zhu Lin” and it is
one of the grottoes in the early period of our country. It
was first built at Northern Liang of the Sixteen States
Period. Then the entension of many generations makes
it more and more magnificent and some scholars
believe that it is the originator of Chinese grottoes.
Three grottoes which are 17 niches, hundreds of square
meters survived murals, more than 100 buddha statues,
many manuscripts in Chinese and Tibetan language of
the Wei Dynasty, Han Dynasty and Tang Dynasty are
preserved. The giant seated Tathagata Buddha statue in
the Great Budda Grotto is especially vivid and
exquisite.
Corresponding author: Jing Xue, master, main research
field: history of design & art. E-mail: jingxue68@163.com.
The historical data showed that: TianTi Mountain
Grotto is the king of the Northern Liang Ju Drainage
Mengxun drilled in the year between 412 and 439. The
grotto is created to dig on the foundation of natural
cave, and already had the history more than 1580 years
till now. In October at the East Jin YuanXi (412), Meng
Sun moved the capital to Guzang from zhangye, and
called it “River West King”. Then he set official bureau,
repaired palace, started to set up city gate. And he
requested eminent monk named Tan Yao in Liang state
and skillful artisans to split mountain to open a new
road in the meantime. They excavated TianTi
Mountain Grotto. Soon his mother was died of illness,
so he digged the status for 5 meters height, who was in
deep desperate on the face meant to repent. The grotto
preserved the handwritten book in North Wei, Sui and
Tang’s period. The early Tang painted the Statue of
bodhisattva on silk. And it conserved the pure statue,
and mural of Tang, Five age, Xixia (Song), Yuan, Qing
etc.. The geography of the TianTi Mountain Grotto is
dangerous. The Buddhism is prosperous when the
culture in Liang periods developed. The excavated hole
caused attention in Buddhism field and the eminent
monks in Western Regions followed. They preached in
 Dazzling Pearl-Analysis of the Artistic Characteristics of Wu Wei’s TianTi Mountain Grotto
the Liang State and translated Buddhist sutra, so TianTi
Mountain Grotto has great reputation [2].
When founding TianTi Mountain Grotto, it
developed many clever artisans and the professional
painters there. After TianTi Mountain Grotto
completed, as the political core and Buddhism center
moved east, many painters also moved east to
Pingcheng (now the Datong in Shanxi province),
starting to excavate statues there. So they developed
Yungang Grottoes, Longmen Grottoes with the
important technical role. However, Northern Wei
Dynasty ended Hexi-the region, which with 140 years
separatist and prosperous situation. The prosperous
Buddhism in Liangzhou and its art value were
subjected to serious wound. Some monks moved to
Liang State, some moved to Pingcheng, and others
moved west to Dunhuang. This was also contributed to
the prosperity of Dunhuang Buddhist. So Dunhuang
became the river west Buddhism center after the
Liangzhou, and pushed river west stone cave culture
development to the second peak within the quick
development of culture in Dunhuang Caves. The art
style of Tang dynasty Buddha in the Dunhuang grottoes
(130 caves) was similar with the Tiantishan Grotto. So
the art style and the invention of the Tianti Grotto
influenced Dunhuang Grotto and Hexi Grotto directly.
From the simple narrative above, we can clearly see
that the effects of the Grotto are: Tianti Mountain
Grotto-Yungang Grotto-Longmen Grotto-Dunhuang
Grotto. It is not intended to diminish the influence of
Mogao Grotto and praise Tiantishan Grotto here. But in
fact, we know the impact of the Tianti Grotto in the
center plain.
3. The Shape Characteristic of Grottoes
Now, only three layers of grottoes and 17 grottos of
different sizes remained in TianTi Mountain. One large
grotto is 30 meters high, 19 meters wide, and 6 meters
deep. There is a large statue of Sakyamuni inside the
grotto, which is 28 meters high and 10 meters wide, it
stands there facing the water, with right arm reaching
129
out and pointing to the front, sitting stably. On both
sides, there are six statues of Manjusri buddisattva,
Samantabhadra, lute heavenly king, Kasyapa and
Ananda, the shapes are vivid and have an air of dignity.
On the northern and southern walls of the grotto were
painted with large murals, the whole murals have clear
strokes, bright color and vivid images. The upper part
of the southern wall is painted with image of blue
dragon with moire-decorated body; the central part is
painted with the images of elephant and deer, the
elephant is packed with shining volumes on the back,
and the lower part is depicted with tiger, tree and
flower. The upper part of the northern wall is painted
with the images of blue dragon and two tigers, the
central part is painted with white horse, black tiger and
Linden, with shining volume on horseback; the lower
part is painted peony flower [3].
3.1 Buddha Statues
As shown in Fig. 1, the Buddha has high bun with
twists and turns, full round face shape, symmetric and
beautiful body, with modest proportion, dressed in
robes through the shoulders, with lotus position, left
hand placed on the belly, palms up, right hand
supporting the knee with fingers down, like a
“monsters felling stamp”. The overall image is plump
Fig. 1 Tang Dynasty, 130 cm high (originally in TianTi
Mountain, 2nd grotto western wall, now preserved in Gansu
Provincial Museum Exhibition Hall).
130 Dazzling Pearl-Analysis of the Artistic Characteristics of Wu Wei’s TianTi Mountain Grotto

and full, easy and sturdy. Drapery turns to become

realistic. From its empty and solemn look and
seemingly sleepy and waking eyes staring downward,
it seems to reveal an eternal mercy and broad mind. So
that the viewers can produce a great feeling of comfort.
Its symmetric and vigorous body and lofty and solemn
look have a high degree of harmony and unity. It fully
shows the high level of Buddhist statues in the Tang
Dynasty. In spite of the moderate rehabilitation by
future generations, it still has very attractive charm.
As shown in Fig. 2, the Buddha has spiral shells bun,
plump face, short neck, wide shoulder, and physical
fitness. It is dressed in semi-shawl robe, in lotus
position, with left hand supporting the knee, right hand
holding clothes, serene and calm. The head is slightly
forward, and eyes are down with free and easy, quiet
and solemn air as if without ego. Its slender eye corners, Fig. 2 Tang Dynasty, 108 cm high (originally in TianTi
solemn eyes and slightly forward body show kind, Mountain Grottoe 3rd grotto, now preserved in Gansu
approachable, solemn and graceful characters, so that Provincial Museum).
the viewers feel amiable and respectful with concerns
arising spontaneously. Ladder style arc drapery is
natural, realistic and rich in decoration, the delicate
skin is like flesh and blood and with texture.
Sophisticated model and the use of modeling
techniques plastic are skilled, reflecting the high level
of statues in the Tang Dynasty. It is one of the most
representative works among statues of the Tang
Dynasty in TianTi Mountain Grotto.

3.2 The Statue of Bodhisattva

As shown in Fig. 3, the Bodhisattva has high bun,

long hair on shoulder, round face, curved eyebrow and
beautiful eyes, well-proportioned body, bare upper
body, lower body with skirt close to legs, shawl
naturally sagging from the shoulder, with the right arm
stretching down, left arm holding in front of chest,
fingers clenched, and the body is standing on circular
lotus stand which is in slightly “S” shape. It has a
charming and soft sense of dynamics, making the Fig. 3 Tang Dynasty, 180 cm high (originally in TianTi
viewers feel calm and not dull, full of vitality in Mountain Grotto 3rd grotto, now preserved in Gansu
serenity. Its dignified posture, subtle look, white jade Provincial Museum).
 Dazzling Pearl-Analysis of the Artistic Characteristics of Wu Wei’s TianTi Mountain Grotto 131

skin and unearthly grace fully reflect the era style of strong lines, while the contours of the skin edge are
statues art in the Tang Dynasty. The natural flow of dyed with relatively wide red lines. Both its statue and
arc-shaped skirt pattern is like cloud and water, with painting technique clearly show the strong influence
“clothing out of water” feature, making its body have of western painting styles. It is valuable material for
more artistic charm due to the perfect air, and the study of early Buddhist paintings of China and
embodying the spirit of the Tang Dynasty. cultural exchange between East and West [4].
As shown in Fig. 4, the Bodhisattva has high bun
with belt, long hair on shoulder, square and round face,
fine eyebrow and big eyes, long eye corners, straight
and high nose, small mouth and thin lips, sturdy and fit
body. The upper body is bare, wearing a collar and
short muktahara, and the lower body wears skirt,
which is close to the legs. The scarf sags from the
shoulders under the arms, and the left hand holds
upward with palm up. A bar is held between thumb and
index finger, the right arm extends downward, with a
garland held in hand, standing loftily on the lotus table.
Analyzed from the garland held in right hand, it may be
the Bodhisattva, one of the assisting Buddha of
Amitabha.
The Bodhisattva body is outlined with “forceful and
continuous” iron wire, dress pattern is dense, lines are
forceful, eyebrows, nose, neck and skin edge are dyed
with white or red color, apparently impacted by uneven
the Western painting. The facial features and body Fig. 4 Bei Liang, about 70 cm (originally in TianTi
features are more close to image of people in western Mountain 4th grotto, now preserved in Gansu Provincial
Museum).
border. Works of two different painting styles of
Chinese Western appear in the same grotto, also in
Bingling Temple 169th Grotto, reflecting the exchange
of Chinese and Western cultures and integration and
tolerance in the early period of Buddhist art in China.

3.3 Worshiping Bodhisattvas

As shown in Fig. 5, the Bodhisattvas has high bun

and long hair, square and round face, fine eyebrows
and big eyes, high and straight nose, with circular
earrings, sturdy body. Upper body is bare, wearing
necklace, armlet and bracelet, bare right shoulder
wearing robe, with both hands held in front of chest
Fig. 5 Bei Liang, about 40 cm (originally in TianTi
folding, kneeling-style worshiping Buddha. The face Mountain 4th grotto central column, now preserved in
reveals a stable look. The profile is outlined with Gansu Provincial Museum).
132 Dazzling Pearl-Analysis of the Artistic Characteristics of Wu Wei’s TianTi Mountain Grotto
4. Conclusions
The TianTi Mountain Grotto reflected the style and
features of our Buddhist art at early period and they are
very valuable to the culture and art exchange between
China and western countries. Although it has been
moved to another place, its value in Chinese Buddhist
art should not be neglected. Its establishment and
development paved the way for the progress of grotto
art of the region of the Hexi Corridor and even the
Central Plain of China. Therefore, the TianTi Mountain
Grotto occupies a very outstanding position in
numerous stone cave temples in our country.
References
[1] L.H. Xue, Wu Wei history: During the reign of Qianlong,
1965, pp. 41-45.
[2] K.G. Zhong, Sweet Cultural Object in the Su Province
Studies of Ancient Relics Graduate School Plait, River
West Stone Cave, Cultural Object Publisher, 1985, pp.
30-31.
[3] L.P. Chen, Y.H. Wang, Chinese Stone Cave Art, Ages Art
Publisher, 2005, pp. 59-60.
[4] D.H. Xu, Statue of Buddha Art Shape, Cultural Publisher,
Shanghai, China, 2008, pp. 67-70.
Editor’s Acknowledgement

All manuscripts submitted to Journal of Environmental Science and

Engineering are usually reviewed by at least one member of the editorial

board members. This commits them to up to 12 assessments each year

and the Editor thanks them for so generously giving their time and advice.

The Editor is also grateful to many other assessors who have helped in

2011.

At the same time, every manuscript submitted to Journal of

Environmental Science and Engineering is refereed by at least 2 outside

experts in the field. The Editor and the Editorial Executive Committee

gratefully acknowledge the considerable time and help the reviewers who

gave during 2011. The Editor thanks very much for their advice and for

their promptness in dealing with the manuscripts.

The Editor also thanks for the understanding, support and help from

our authors, readers and other people and agencies.