Prof

Journal of Agricultural Engineering and Technology (JAET), Volume 22 (No.2) June, 2014

NIGERIAN INSTITUTION OF AGRICULTURAL ENGINEERS National Executive Committee 2013 S/N Name Position 1 Engr. (Prof.) B. A. Adewumi National Chairman 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

19 20 21 22

E-mail address [email protected] or [email protected] Engr. (Dr.) S. M. Musa National Vice- Chairman [email protected] Engr. (Prof.) S. Z. Abubakar Immediate Pass Chairman [email protected] Engr. (Dr.) M. A. Enaboifo National Secretary [email protected] Engr. (Dr.) J. C. Adama Assistant National Secretary [email protected] Engr. B. G. Jahun National Financial Secretary [email protected] Engr. (Mrs.) C. A. Adamade National Treasurer [email protected] Engr. (Dr.) M. K. Othman Publication Secretary [email protected] Engr. S. Bello National PRO [email protected] or [email protected] Engr. (Dr.) S. A. Iya North East Regional Chairman [email protected] Engr. (Dr.) L. A. O. South West Regional Chairman [email protected] Ogunjimi Engr. A. A. Akintola North Central Regional [email protected] Chairman Engr. A. L. Ijasan South South Regional Chairman [email protected] Engr. C. C. Emekoma South East Regional Chairman [email protected] or [email protected] Engr. J. O. Daudu North West Regional Chairman [email protected] Engr. Y. K. Dalha Ex. Officio [email protected] Engr. (Mrs.) N. I. Nwagugu Ex. Officio [email protected] Engr. I. Azogu Executive Director, National [email protected] Centre for Agricultural Mechanization, (NCAM) Eng. M. Y. Kasali Business Manager [email protected] Engr. (Dr.) F. Alonge Auditor 1 [email protected] Engr. (Mrs.) R. S. Samaila Auditor 2 [email protected] Engr. (Prof.) A. P. Onwualu Editor-in-Chief, JAET [email protected]

Nigerian Institution of Agricultural Engineers © www.niae.net

i


JOURNAL OF AGRICULTURAL ENGINEERING AND TECHNOLOGY (JAET) EDITORIAL BOARD Editor-In-Chief Professor A. P. Onwualu, FAS National University Commission (NUC) 26, Aguiyi Ironsi Street, Maitama District, Abuja, Nigeria [email protected]; [email protected] Phone: 08037432497 Prof. B. Umar – Editor, Power and Machinery Rector, Adamawa State Polytechnic, Yola, Adamawa State, Nigeria. E-mail: [email protected] Phone: 08023825894 Prof. A. A. Olufayo – Editor, Soil and Water Engineering Agricultural Engineering Department, Federal University of Technology, Akure, Ondo State, Nigeria. E-mail: [email protected] Phone: 08034708846 Prof. A. Ajisegiri – Editor, Food Engineering College of Engineering, University of Agriculture, Abeokuta, Ogun State, Nigeria. E-mail: [email protected] Phone: 08072766472 Prof. K. Oje – Editor, Processing and Post Harvest Engineering Agric. and Bio-resources Engineering Department, University of Ilorin, Kwara State, Nigeria. E-mail: [email protected] Phone: 08033853895 Dr. A. El-Okene – Editor, Structures and Environmental Control Engineering Agricultural Engineering Department, Ahmadu Bello University, Zaria, Kaduna State, Nigeria. E-mail: [email protected] Phone: 08023633464 Prof. D. S. Zibokere – Editor, Environmental Engineering Agric. and Environmental Engineering Dept., Niger Delta University, Wilberforce Island, Yenegoa. E-mail: [email protected] Phone: 08037079321 Prof. C. C. Mbajiorgu – Editor, Emerging Technologies Agricultural and Bioresources Engineering Department, University of Nigeria, Nsukka, Nigeria. E-mail: [email protected] Phone: 07038680071 Prof. (Mrs) Z. D. Osunde – Editor, Processing and Post Harvest Engineering Agricultural Engineering Department, Federal University of Technology, Minna, Niger State, Nigeria. E-mail: [email protected] Phone: 08034537068 Mr. Y. Kasali – Business Manager National Centre for Agricultural Mechanization, PMB 1525, Ilorin, Kwara State, Nigeria. E-mail: [email protected] Phone: 08033964055 Dr. J. C. Adama – Editor, Agricultural Mechanization Agricultural Engineering Department, University of Agriculture, Umudike, Abia State, Nigeria. E-mail: [email protected] Phone: 08052806052 Dr. B. O. Ugwuishiwu – Editor, Farm Structures and Waste Management Agricultural and Bioresource Engineering Department, University of Nigeria, Nsukka, Nigeria. E-mail: [email protected] Phone: 08043119327 Miss I. C. Olife – Technical Assistant to Editor-In-Chief Raw Materials Research and Development Council, Abuja, Nigeria. E-mail: [email protected] Phone: 08033916555 Nigerian Institution of Agricultural Engineers © www.niae.net

ii


Aims and Scope The main aim of the Journal of Agricultural Engineering and Technology (JAET) is to provide a medium for dissemination of high quality Technical and Scientific information emanating from research on Engineering for Agriculture. This, it is hoped will encourage researchers in the area to continue to develop cutting edge technologies for solving the numerous engineering problems facing agriculture in the third world in particular and the world in general. The Journal publishes original research papers, review articles, technical notes and book reviews in Agricultural Engineering and related subjects. Key areas covered by the journal are: Agricultural Power and Machinery; Agricultural Process Engineering; Food Engineering; Post-Harvest Engineering; Soil and Water Engineering; Environmental Engineering; Agricultural Structures and Environmental Control; Waste Management; Aquacultural Engineering; Animal Production Engineering and the Emerging Technology Areas of Information and Communications Technology (ICT) Applications, Computer Based Simulation, Instrumentation and Process Control, CAD/CAM Systems, Biotechnology, Biological Engineering, Biosystems Engineering, Bioresources Engineering, Nanotechnology and Renewable Energy. The journal also considers relevant manuscripts from related disciplines such as other fields of Engineering, Food Science and Technology, Physical Sciences, Agriculture and Environmental Sciences. The Journal is published by the Nigerian Institution of Agricultural Engineers (NIAE), A Division of Nigerian Society of Engineers (NSE). The Editorial Board and NIAE wish to make it clear that statements or views expressed in papers published in this journal are those of the authors and no responsibility is assumed for the accuracy of such statements or views. In the interest of factual recording, occasional reference to manufacturers, trade names and proprietary products may be inevitable. No endorsement of a named product is intended nor is any criticism implied of similar products that are not mentioned. Submission of an article for publication implies that it has not been previously published and is not being considered for publication elsewhere. The Journal’s peer review policy demands that at least two reviewers give positive recommendations before the paper is accepted for publication. Prospective authors are advised to consult the Guide for Authors which is available in each volume of the Journal. Four copies of the manuscript should be sent to: The Editor-In-Chief Journal of Agricultural Engineering and Technology (JAET) ℅ The Editorial Office National Centre for Agricultural Mechanization (NCAM) P.M.B. 1525 Ilorin, Kwara State, Nigeria. Papers can also be submitted directly to the Editor-In-Chief or any of the Sectional Editors. Those who have access to the internet can submit electronically as an attached file in MS Word to [email protected]; [email protected]. All correspondence with respect to status of manuscript should be sent to the Technical Assistant to the Editor-In-Chief at [email protected].


iii


TABLE OF CONTENTS NIAE National Executive Committee 2013

…

…

…

i

Editorial Board

…

…

…

…

…

ii

Aims and Objectives

…

…

…

…

…

iii

Table of Contents

…

…

…

…

…

iv

…

…

1

Development of a Motorized ‘Egusi’ Melon Seeds Oil Expeller R. S. Samaila and O. Chukwu … …

…

…

13

Development of a Tomato Slicing Machine E. C. Oriaku, C. N. Agulana and N. I. Nwagugu

…

…

…

26

Palm Kernel Oil Expression by Uniaxial Compression I. C. Ozumba and K. Oje …

…

…

…

37

Development of a Solar Cabinet Dryer for Root Crops Chips in Nigeria C. O. Nwajinka and C. U. Onuegbu … … …

...

47

Effect of Drying Methods and Production Process on the Quality Parameters of Unfermented Cassava Flour C. A. Adamade, B.A Jackson and F. Agaja … …

…

59

Development of a Portable Air Flow Digital Meter for Grain Drying A. B. Istifanus and C. C. Mbajiorgu … …

…

65

Assessment of Problems Associated with Tractor Hiring Services in Kura and Garun-Mallam Local Government, Kano River Irrigation Project M. S. Abubakar, I. Lawan and A. A. Wazeer

…

…

Characterization and Disaggregation of Daily Rainfall Data of Onitsha, Anambra State, Nigeria G. I. Ezenne, H. J. Ugwuozor and C. C. Mbajiorgu … … …

76

Comparison of ACRU and HEC-HMS Models in Runoff Prediction in a Watershed, South West Nigeria A. P. Adegede, K. N. Ogbu, V. Ogwo and C. C. Mbajiorgu …

…

88

Infiltration through Traffic Compacted Soil I. E. Ahaneku … …

…

97

Investigation of Ground Water Quality during the Dry Season in Rivers State, Nigeria: A Case Study of Port Harcourt Metropolis I. Fubara-Manuel and R. B. Jumbo … … …

…

105

Modeling Channel Roughness Coefficient using Dimensional Analysis M. Y. Kasali and A. O. Ogunlela … …

…

112


…

…

…

iv


Physico-Chemical and Strength Properties of Some Local Materials for Irrigation Canal Lining M. Y. Kasali, A. O. Ogunlela, D. James and I. I. Azogu …

…

121

Water Temperature Simulation of Owena River at Owena Dam Site Using HEC-RAS Model G. I. Ezenne, E. L. Ndulue and O. L. Yakubu … … …

131

Guide for Authors

…

…


…

…

…

140

v


ASSESSMENT OF PROBLEMS ASSOCIATED WITH TRACTOR HIRING SERVICES IN KURA AND GARUN-MALLAM LOCAL GOVERNMENT, KANO RIVER IRRIGATION PROJECT M. S. Abubakar, I. Lawan and A. A. Wazeer Department of Agricultural Engineering, Bayero University, Kano, Nigeria. Email: [email protected] ABSTRACT A study was undertaken to assess the tractor hiring services in Kura and Garun-Mallam local government areas of Kano state. The criteria used were fleet of equipment, personnel, workshop facilities, tractors/implement (make and model) and its impact to farming populace. The study adopted the use of questionnaire and personal interviews. Amongst the information requested from the respondents were maintenance problems, common breakdown, level of experience of operator/mechanics, annual expenses on maintenance, annual gains from hiring, cost of operation, availability of the tractors, problems encountered and other expenses made by the farmer apart from operational charges. Recommendations were presented towards achieving this goal which include; adoption of regular maintenance culture, staff recruitment, and retraining and motivation of operators as ways of improving the success of the peasant farmers, private and public sectors respectively. KEYWORDS: Tractor hiring services, irrigation, agricultural mechanization. 1.

INTRODUCTION

Farm power may be described as any source of energy that makes power available for farming operations, tools and implements and powered machinery for agricultural production. Odigbo (2000) defined mechanization as use of machinery, any machine to accomplish a task or an operation involved in agricultural production. This differentiates mechanization from tractorization which according to Azogu (2009), simply means the use of tractors to provide farm power for carrying out farm work. The farm operation in Kura and Garun-Malam basically includes; tillage activities such as ploughing, harrowing, ridging and dyke making. Tractorization therefore forms an integral part of mechanization. Farm power makes power available to farming operation while farm machinery is a collection of machines for farm operation and includes all types of implements and devices such as plough, harrow, tractors etc. Tractor is the most expensive item of all farm machineries. As a result of high cost of tractors and implements, ownership which is mainly by the government through the ministries of agriculture and lately by the departments and parastatals whose mandate make service available to the farming populace who are not economically strong to acquire equipment, tractors and implements. Special units were established and referred to as Tractor Hiring Unit (Mijinyawa and Kisaiku, 2006). Tractor ownership is the process by which an organization which is mainly government owned through ministry of agriculture, departments and private sectors own tractors, implements and equipments for the purpose of agricultural production render service to faming populace who cannot afford to buy tractors, implements and equipments and also to make business for profit. The Community Cooperative Tractor Services (CCTS) through PPP was initiated in 2008. As a test case and in line with the 2008 budgetary provision a total of 1,950 units of tractors and implements were provided to 16 principal supervisor officers (PSOs) under PPP, as incentive and to assist the PSOs to jumpstart the exercise, the Federal Government released 25% grant to them upon submission of advance payment guaranteed from their bankers. Unfortunately however, 6 months after the expiration of the delivery period, only about 30% success could be recorded and based on complains received from the PSOs and field observations by the NFRA staff, the list of the PSOs was reviewed and alternative strategy adopted (Elesa, 2010). Nigerian Institution of Agricultural Engineers © www.niae.net

1


Tractor Hiring Services (THS) can be define as service which evolved as a result of high cost of tractors, implements, equipments ownership which is mainly by the government through the Ministry of Agriculture and lately by departments and private sectors so as to make service available to the farming populace who are not economically strong to acquire the tractors, implements and equipment (Mijinyawa and Kisaiku, 2006). The Major Stakeholders of the Tractor Hiring Service through the PPP include: i. The Federal Government represented by the Federal Ministry of Agriculture and Water Resources. Hadejia Jama’are River Basic Development Authority, Federal Ministry of Agriculture, The Federal Government has its outlets for Kura and Garun-Mallam. ii. The State and/or Local Government who opt to participate have their outlets in Kano State Ministry of Agriculture, Kano State Agricultural and Rural Development Authority iii. The Local Government has its outlet in the Departments of Agriculture in the Local Governments (Elesa 2010). The agreement involved in tractor hiring services includes the following: i) Payment for the service delivery ii) Charges involved in tractor transportation to the field of operation by the public sectors. iii) Minor repairs iv) Fueling of the tractors Agricultural development in many states of Nigeria depends on how successful these tractor hiring services are. It is therefore important to assess how the services are fairing in order to identify problems and proffer solutions. The objective of this work was to assess the performance of tractor hiring services in Kura and Garun-Mallam Local Government Areas of Kano State, Nigeria.

2.

METHODOLOGY

2.1

Study Area

Kura is located 40 km from Kano city at longitude 80 51’ and latitude 110 98’ (Kano Census, 2006). 2.2

Data Collection

Data was collected from both private and public establishments using a questionnaire and personal interviews. 2.2.1 Questionnaire Design and Administration Sixty questionnaires were administered to the farmers, tractor operators, mechanics, private and public sectors for the purpose of information collection in the study area. 2.2.2 Personal Interview Additional information was equally gathered through personal interviews with the private tractor owners, public sectors such as Ministries and Zonal Offices, and staff of the tractor hiring units.

3.

RESULT AND DISCUSSION

3.1

Source of Capital to Purchase Tractors by the Private Sector Operators

Figure 1 indicates that 53.33% of the respondents source their capital through personal savings, while 40% of the respondents obtained their own through PPP, and 3.33% through bank loans and others. Nigerian Institution of Agricultural Engineers © www.niae.net

2


Therefore, this shows that the Federal Government intervention through PPP and support from banks by giving loans is still very negligible. 80 70 60 50 40

Percentage (%)

30

Frequency

20 10 0 Personal savings

PPP

Bank loan

Others

Figure 1: Sources of Capital 3.2 Functionality of the Tractors Figure 2 shows the results obtained for the number of tractors that are functional, have minor repairs, major repairs or scrap(s), 33.33% of the tractors in Kura and Garun-Mallam Local Government are functional, 16.67% have minor repairs, 21.67% have major repairs and lastly 28.33% of the tractors are scrap. Therefore, this shows that majority of the tractors in the study area are not functional. 60 50 40 Percentage (%)

30

Frequency

20 10 0 Functional

Minor repairs

Major repairs

Scrap

Figure 2: Functionality of Tractors 3.3 Tractor Service Schedule Figure 3 shows that 20% of the tractors and implements in the study area are serviced every month, 13.33% are serviced every six months, 23.33% in a year, 1.67% in two years no matter the hours spent in the field. 41.67% of the tractors undergo maintenance in a variable time. This revealed that there is no form of regular maintenance schedule usually followed by the stake holders in the study area.


3


60 50 40 Percentage (%)

30

Frequency

20 10 0 6 months

1 year

2 years

Others

Figure 3: Tractor Service Schedule 3.4 Tractor Overhauling Schedule Figure 4 shows that 25% of the tractors are overhauled every 6 months, 35% in one year, and 13.33% in two years. 26.66% of the tractors are overhauled in variable time. Therefore, this revealed that there is gross misuse of the tractors in the study area as up to 25% usually require overhauling in less than a year. 60 50 40 Percentage (%)

30

Frequency

20 10 0 6 months

1 year

2 years

Others

Figure 4: Tractor Overhauling Schedule 3.5

Annual Cost of Tractor Maintenance

Figure 5 shows that 26.67% of the respondents spent below N70,000.00, 48.33% of the respondents spent within the range of N70,000.00 – N500,000.00, 15% of the respondents spent in the range of N500,000.00 – N1000,000.00 while 10% respondents spent above N1,000,000.00.


4


90 80 70 60 50 40 30 20 10 0

Percentage (%) Frequency

Below #70,000

#70,000 – #500,000

#500,000 – #1000,000

Above #1000,000

Figure 5: Cost of Tractor Maintenance 3.6

Annual Gains of the Tractor Hiring Service

Figure 6 shows that 13.33% of the respondents gained below N100,000.00, 30% of the respondents gained in the range of N100,000.00 to N500,000.00, 33.33% of the respondents gained in the range N500,000.00 to N1000,000.00 and 23.33% of the respondents gained above N1,000,000.00. Therefore, this shows that an average respondent gained within the range of N500, 000.00 – N1, 000,000.00. 60 50 40 30

Percentage (%)

20

Frequency

10 0 #Below 100,000

#100,000 – #500,000

#500,000 – #1000,000

Above #1000,000

Figure 6: Annual Gain of Tractor Hiring Service 3.7

Age of Tractors

Figure 7 shows the range of year(s) of the tractors in the study area. It shows that 10% of the tractors have age below one year, 26.67% of the tractors fall in the range between 1 to 5 year(s), 50% of the tractors fall in the range between 5 – 10 years, while 13.33% of the tractors are in the range above 10 years. This indicated that on the average the tractors in the study area are within the range of 5 to 10 years.


5


90 80 70 60 50 40 30 20 10 0


Below 1 year

1 – 5yrs

5 – 10yrs

Above 10yrs

Figure 7: Age of the Tractors 3.8

Level of Formal Education of the Tractor Operators

Figure 8 indicates the findings on the educational status of the operators. As reported by Shitu (2011), formal education is one of the factors that affect the capability of tractor operators. Therefore, it is one of the keys in assessing his management ability of the machine/implements due to his direct contact with the machine/implement. The level of formal education status of a farm operator will influence his understanding the instruction on the study manual of the tractor/implement. As it is stated that illiterates are limited in their understanding of the manufacturer manual (Shitu, 2011). 90 80 70 60 50

Percentage (%)

40

Frequency

30 20 10 0 Primary

Secondary

Tertiary

None

Figure 8: Level of Formal Education of the Operators From the result obtained, it was observed that 25% of the operators have primary certificate, 16.67% of the operators are with secondary certificate, 10% of the operators obtained tertiary certificate and up to 48.33% of the operators have not attended any form of formal education. This revealed that the most of the operators in the study area have low level of western education. 3.9

Experience of Tractor Operators

Figure 9 shows that 40% of the tractor operators have experience of range between 1 to 10 years, 30% have experience between 10 – 20 years, 20% have experience in the range of 20 – 40 years, while 10% have experience of 40 years and above.


6


70 60 50 40

Percentage (%)

30

Frequency

20 10 0 1 – 10yrs

10 – 20yrs

20 – 40 yrs

Above 40yrs

Figure 9: Tractor Operators Experience 3.10 Hours Spent During Field Operation Per Day Figure 10 shows that 3.33% of the respondents claimed that they spent below 5 hours per day, 60% of the respondents spent 5 – 8 hours, 30% of the respondents spent 8 – 10 hours, and while 6.67% of the respondents spent above 10 years. 120 100 80 Percentage (%)

60

Frequency

40 20 0 Below 5hr

5 – 8hr

8 – 10hr

Above 10hr

Figure 10: Hours Spent in the Field 3.11 Rate of Tractor Breakdown Figure 11 shows the frequency of breakdown of the tractors in the study area. Results revealed that 28.33% of tractors usually breakdown in 1 – 6 months, 25% of the tractors breakdown in 6 months to 1 year and 46.67% of tractor brake down in variable occurrence.


7


80 70 60 50 40

Percentage (%)

30

Frequency

20 10 0 1 – 6 months

6 months – 1 year

Variable

Figure 11: Rate of Tractor Breakdown 3.12 Experience of the Mechanics Figure 12 at the appendix shows that 28.33% of the mechanics have experience of below five years, 40% have experience of between 5 5o 10 years, and 16.67% have experience of between 10 – 25 years while 15% have their level of experience above 25 years. 70 60 50 40

Percentage (%)

30

Frequency

20 10 0 1 – 10yrs

10 – 20yrs

20 – 40 yrs

Above 40yrs

Figure 12: Experience of the Mechanics 3.13 Availability of Spare Parts Figure 13 shows that 5% of the spare parts are obtained from public agro stores, 80% from open market and 15% from private stores. This shows that 80% of the spare parts can be obtained from open market.


8


90 80 70 60 50 40 30 20 10 0


Poor

Good

Very good

Figure 13: Availability of Spare Parts 3.14 Maintenance Practice Performed on Tractors and Implements Figure 14 shows that 16.67% of the respondents perform daily maintenance, 50% perform corrective maintenance, 12% carryout preventive maintenance and while 13.33% carried out all the types of maintenance. 90 80 70 60 50

Percentage (%)

40

Frequency

30 20 10 0 Daily

Corrective

Preventive

All of the above

Figure 14: Maintenance System Practice 3.15 Sources of Tractors Used Figure 15 shows that 40% of the farmers got their tractors from private organizations, 25% from the public sectors, while 35% obtained theirs from both private and public sectors.


9


140 120 100 80

Percentage (%)

60

Frequency

40 20 0 Public agro stores

Open market

Private stores

Figure 15: Sources of Tractor Used 3.16 Availability of the Tractors to the Farmers Figure 16 shows that 10% of the respondent farmers complained of the poor availability of the tractors, 30% of the respondents claimed that the tractor availability is fairly good, 40% responded that it is good, while 20% claimed that the tractor availability is very good. 70 60 50 40

Percentage (%)

30

Frequency

20 10 0 Poor

Fairly good

Good

Very good

Figure 16: Availability of Tractors 3.17 Charges of Tractor Hiring Service Table 1.0 shows the amount of money the farmers are charged per operation in an acre. From the result obtained, it can be seen that charges for the operations carried out by different stakeholders vary. By making physical analysis of the result, the State Government charges is lower than that of the Local Government charges. In Kura and Garun-Mallam Local Government Areas, the private organizations providing the services of tractor hiring charged higher than the State and Local Governments.


10


Table 1.0 Cost of operation the farmers are charged per acre S/NO Type of operation Charges by state Local Government charges 1 Harrow (single) 1000 1500 2 Harrow (double) 2000 3000 3 Ploughing 3500 4000 4 Ridging 1500 2000 5 Dyke making 1500 2000 6 Trailing 3000/day 4000/day 7 Slashing 10,000 -

Private charges

sectors

2000 – 3000 4000 – 6000 6000 and above 2500 – 3000 3000 5000/day -

3.18 Expenses Paid Apart from the Service Charges The expenses paid by the peasant farmer apart from the service charges included the following based on agreements:  Overtime charges According to the farmers, there is additional money they pay to the operators if the hours to perform the operation exceeded as the agreement were made.  Fueling charges Fueling of the tractors is also part of the extra charges paid for transporting the tractors from their basement to the field of operation. These changes are normally collected by the public hiring units and sometimes even by the private organizations.  Minor repairs charges This is another problem encountered during the operation which were not part of the agreement made between the farmers and the tractor operators are mostly charged for minor repair(s). The farmers stated that they pay mostly for this minor repairs to avoid time wasting that may affect the period of production and also to prevent missing opportunity of getting the tractors. 4.

CONCLUSION AND RECOMMENDATIONS

4.1

Conclusion

Analysis of the result obtained from the study area revealed that 53.33% of the private sectors obtained their tractors/implements through personal saving and amongst the tractors found only 33.33% were found to be functional but the remaining percentages have problems of minor repairs, major repairs and scraps. The time duration of tractor usage is in the range of 5 – 10 years. This duration shows downtime of the tractor lifespan and its ability to likely have problem. It was also found that 40% operators/mechanics have low-level of formal education which indicates their limitations to the understanding of the tractor user’s manual. 40% of the breakdowns that occur are breakage of implements and bearing from the result analyzed in the case study area. This problem occurs normally due to lack of preventive maintenance and misuse of the machine by the operators, poor knowledge of detecting problems of the machine due to illiteracy. It was also found that the farmer’s profit is affected by minor repairs he paid for during operation, high charges charged by the private sectors, variability on the amount paid for overtime, transportation of the tractors to the field of operation and so on. Additional information gathered through personal interviews while administering the questionnaire revealed the problem of the use of middle men (“Yankamisho”) which increases expences to both the farmer and the tractors owners. 4.2

Recommendations

In respect to the various findings from the information gathered, hopefully the areas of weaknesses could be addressed with adequate implementation of the following recommendations: Nigerian Institution of Agricultural Engineers © www.niae.net

11


i. ii.

iii. iv. v.

Government at all levels must provide support for both cooperative and individuals for owning tractors at subsidized rate. Governmental and non-governmental organizations (NGOs) should be organizing training workshops to sensitize tractor owners, operators, farmers and tractor mechanics on effective tractor maintenance strategy. To adopt regular maintenance culture since it is cheaper and prolongs lifespan of the tractor and implements. Proper training on tractor and implement repairs, adjustable/replacement should be provided to both tractor operators and mechanics to avoid tractor breakdown. There is need for the hiring services providers to give emphasis on proper organization, planning, utilization and management of the tractor hiring units. There is need for the Government to come up with a policy that will be providing farmers with hiring services at a subsidized rate to support them.

REFERENCES Adoga, G.A., Thomas M.T. 2013. Tractor/machinery hire service in Kaduna state: Overview, challenges and way forward. Department of Agric. Engineering service, Ministry of Agriculture Kaduna, Kaduna State. Azogu, I.I. 2009. Promoting appropriate mechanization technologies for improved agricultural productivity in Nigeria: the role of the national center for Agricultural Mechanization – Journal of Agricultural Engineering Technology (JAET) Vol. 17, No. 2 Ahmed H.I. 2006. Predictive maintenance of tractors for higher utilization in Kano state of Nigeria. Proceedings of the Nigerian Institute of Agricultural Engineers, 28:98 – 105. Mijinyawa, Y. and Kisaiku, O. 2006. Assessment of the Edo State of Nigerian tractor hiring services. Agricultural Engineering International: the CIGR E journal invited overview paper no.10.vol.VIII. Odigboh, F.U. 2000. Mechanization of the Nigerian agricultural industry, Pertinent notes, pressing issues, Pragmatic options. A public lecture delivered at the Nigerian Academy of science, International Conference Center, Abuja, April 15, 2000. Shitu A. 2011. Investigation into failure of some selected tillage implement, case study of Kano State. Unpublished B.Eng. project submitted to the Agricultural Engineering Dept., Bayero University Kano.


12


DEVELOPMENT OF A MOTORIZED ‘EGUSI’ MELON SEEDS OIL EXPELLER R. S. Samaila1 and O. Chukwu2 National Centre for Agricultural Mechanization P M B 1525. llorin, Kwara State, Nigeria. Email: [email protected] 2 Department of Agricultural Engineering Federal University of Technology Minna, Niger State, Nigeria. Email: [email protected]

1

ABSTRACT A motorized melon oil expeller was designed, fabricated and tested. The materials for fabrication were sourced locally. It was developed and fabricated to remove drudgery and reduce losses associated with processing of melon seeds into oil. The machine is simple and can easily be operated and maintained by rural women who are major processors of melon oil. It was tested for throughput capacity (Tp), percent oil yield (Yi) and expression efficiency (Ee). Throughput capacity of 10.12 kg/hr, oil yield of 30.23L/kg and extraction efficiency of 60.83% were obtained when 5 kg of melon seeds were fed into the machine. KEYWORDS: Oil extraction, melon (Egusi) seeds, motorized expeller, machine design. 1.

INTRODUCTION

Melon Citrullus Lanatus SSP Coloeynthoides is an annual herbaceous climber of the family cucurbitaceous. There are different types which vary in size and shape, they are round or oblong. The seeds colour varies, white through brown to black and not all melon flesh and seeds are edible. In Nigeria, the existence of melon dates back to the 17th century (Douglas, 1982). Melon Egusi is a popular fruit in Nigeria because of the edible seeds which are commonly used in the preparation of local soup or stew and snacks such as fried melon seed ball known as “Robo” in South - Western Nigeria. In the East, the seeds are sometimes roosted and eaten as snacks. The seeds are rich in oil (30 – 50 percent) which is comparable to other oil plants (Omidiji, 1997), the oil contains is a high level saturated fatty acids (Adeniran et al, 1981). It is also an important component of the traditional cropping system usually interplant with such staple crops as cassava, maize, sorghum, etc. (Omidiji et al., 1985). There are two basic seed types of egusi melon distinguishable by the presence or absence of seed edges; they are usually stored at a moisture content of 6-7.7%. The first type known locally as “Bara” has thick uniform edge which is either black or white. The second type known as ‘Serewe’ has no distinct edge. The two seeds are differently distributed in the country. The ‘bara’ is most commonly found in the South-Western and Northern part while ‘serewe’ is found in some part of Northwest, East and middle belt of Nigeria. The difference in distribution appears to be based on consumer preference rather than physiological adaptation of the crop (NIHORT, 1976 – 1986). Odigboh (1979) described the unshelled melon seed as very small having a mean major diameter of about 12mm, intermediate diameter of 8mm, a minor diameter of 2.3mm and weighing about 150mg on the average. The advent of new processing technologies has led to different types of oil expelling machines ranging from small hydraulic hand pressing to complex shaft screw press for expelling oil from various oil seeds. However, some of the available extractor/expeller cannot be used for expelling oil from shelled melon seeds because of the peculiar nature of the seed; especially the shape and size. With increasing cultivation


13


of melon it becomes necessary to improve on the existing technology in order to increase the quality and the quantity of oil yield. Therefore, there is the need to improve on the existing expeller by developing shelled melon seeds oil expeller that can be affordable by the rural farmers, and to remove the wastefulness, unhygienic and drudgery involved in manual kneading of the seeds for oil by farmers, thereby boosting their income. The objective of the research was to develop a melon seeds oil expeller. 2.

MATERIALS AND METHODS

2.1

Description and Working Principles of the Melon Oil Expeller

The expeller consists of a prime mover, pulleys and belt drive, happer, extracting chamber, frame and discharge chamber (Fig. 1). The different parts are described below. Prime Mover: This generates power to be transmitted through belt to the expelling unit and produces sufficient torque at the unit. A 1hp, 1420 rpm electric motor was used. Pulleys and Belt Drive: The component transmits power from the prime mover to the expeller shaft. Vgroove pulleys and belts were selected due to their advantage over flat belts. Hopper: This is cylindrical in shape; it serves as the inlet through which the melon seeds are fed into the extracting chamber. The hopper is fabricated from 2 mm thick galvanized iron metal sheet. The Extracting Chamber: This is where the compression of melon seeds is carried out for release of the oil. The extracting unit consists of a cylindrical barrel of 2 mm thick mild steel and helical screw that serves as a conveyor driven by a belt drive of 1hp, 1420 rpm and 3-phase electric motor. . The Heater Band: It is located on the choke cone to heat the melon seeds as it is being compressed. The heater band allows the oil within the cells to come to the surface for easy expression of oil from the seeds. The Discharge Outlet: This is the point where the oil is collected. It is located at the bottom of the barrel. The Frame: This is the mounting support for all the components of the machine.


14


Fig. 1. Exploded View of the Melon Oil Seeds Expeller 2.2

Design Calculations

The following were considered in the design and development of the oil expeller: Efficiency, cost of construction, operating condition, availability of materials and maintenance 2.2.1 Hopper Design The design considerations for the construction of the hopper of the expeller include: (i) The volumetric capacity (ii) The intake mass It is desired that the hopper whose cross-section is shown in Figure 2.0 will handle a maximum of 5 kg of melon seeds. The following equations were used to estimate the required dimensions. Top Cylindrical Part: 𝑽𝒐𝒍𝒖𝒎𝒆, 𝑽𝟏 = 𝝅𝑹𝟏 𝟐 𝒉𝟏

(1)

Where: V1 = volume of the top cylindrical part, π = pie, R1 = radius of the top cylindrical part, h1 = height of the top cylindrical part


15


Fig. 2. Cross-section of the Hopper Conical Part: 𝟏

𝐕 = 𝟑 𝛑𝐡𝟐 ( 𝐑 𝟐 𝟐 + 𝐑 𝟐 𝐫𝟐 + 𝐫𝟐 𝟐 )

(2)

Where: V =Volume bottom cylinder; R2 = radius of conical part = 0.1135 m; r2 = radius of bottom cylinder = 0.0195 m; h2 = height of conical part = 0.148 m; V2 = 0.002398823787 𝑽𝒐𝒍𝒖𝒎𝒆, 𝑽𝟑 = 𝝅𝒓𝟐 𝟐 𝐡𝟑

(3)

Where: r2 = radius of bottom cylindrical part = 0.0195 m; h3 = height of bottom cylindrical part = 0.062 m; 𝐕 = 𝟎 ∙ 𝟎𝟎𝟎𝟎𝟕𝟒𝟎𝟕𝟒𝟐𝟐𝟏𝐦𝟑 Total Volume = V1+V2+V3 = 0.004497m3 Assuming maximum volume of seeds to be processed at a time is 𝟎. 𝟎𝟎𝟒𝟒𝟗𝟕𝒎𝟑 Mass Capacity = 𝝆𝒗 Where: 𝛒 = 𝐝𝐞𝐧𝐬𝐢𝐭𝐲 and 𝐯 = 𝐯𝐨𝐥𝐮𝐦𝐞 𝐁𝐮𝐥𝐤 𝐝𝐞𝐧𝐬𝐢𝐭𝐲𝐟𝐨𝐟 𝐦𝐞𝐥𝐨𝐧 = 𝟔𝟖𝟎 𝐤𝐠⁄𝐦𝟑 (Isiaka et al., 2006)


(4)

16


𝐌𝐚𝐬𝐬 𝐜𝐚𝐩𝐚𝐜𝐢𝐭𝐲 = 𝟔𝟖𝟎 × 𝟎. 𝟎𝟎𝟒𝟒𝟗𝟕 = 𝟑. 𝟎𝟓𝟕𝟗𝟔 𝐤𝐠 𝐌𝐚𝐬𝐬 𝐜𝐚𝐩𝐚𝐜𝐢𝐭𝐲 = 𝟑. 𝟎𝟓𝟖 𝐤𝐠 2.2.2 Determination of Speed Required to Expel Oil from Melon Seeds The ratio of velocities of the driver and the driven is determined by the ratio of the gears that would give the required speed and would expel oil from the melon seeds. The arrangement of the gears is shown in Figure 3.

d1=50

d2 =160 d2’ =70

d3 = 160

Fig. 3. Arrangement of the Gears in the Oil Expeller The ratio of velocities of the driver and the driven is expressed mathematically as 𝒅𝟏 𝑵𝟏 = 𝒅𝟐 𝑵𝟐

(5)

𝐝𝟏 = 𝟓𝟎 𝐦𝐦 𝐝𝐢𝐚𝐦𝐞𝐭𝐞𝐫 𝐨𝐟 𝐭𝐡𝐞 𝟏𝐬𝐭 𝐠𝐞𝐚𝐫, 𝐝𝟐 = 𝟏𝟔𝟎 𝐦𝐦 𝐝𝐢𝐚𝐦𝐞𝐭𝐞𝐫 𝐨𝐟 𝟐𝐧𝐝 𝐠𝐞𝐚𝐫 𝐝𝟐′ = 𝟕𝟎 𝐦𝐦 𝐝𝐢𝐚𝐦𝐞𝐭𝐞𝐫 𝐨𝐟 𝟑𝐫𝐝 𝐠𝐞𝐚𝐫, 𝐝𝟑 = 𝟏𝟔𝟎 𝐦𝐦 𝐝𝐢𝐚𝐦𝐞𝐭𝐞𝐫 𝐨𝐟 𝟒𝐭𝐡 𝐠𝐞𝐚𝐫. 𝐍𝟏 = 𝐬𝐩𝐞𝐞𝐝 𝐨𝐟 𝟏𝐬𝐭 𝐠𝐞𝐚𝐫, 𝐍𝟐 = 𝐬𝐩𝐞𝐞𝐝 𝐨𝐟 𝟐𝐧𝐝 𝐠𝐞𝐚𝐫 , 𝐍𝟐′ = 𝐬𝐩𝐞𝐞𝐝 𝐨𝐟 𝟑𝐫𝐝 𝐠𝐞𝐚𝐫 𝐍𝟑 = 𝐬𝐩𝐞𝐞𝐝 𝐨𝐟 𝟒𝐭𝐡 𝐠𝐞𝐚𝐫 𝑵

𝒅

𝑽𝑹 = 𝑵𝟏 = 𝒅𝟐 𝟐

(6)

𝟏

𝐫𝐞𝐯 𝐬𝐩𝐞𝐞𝐝 𝐨𝐟 𝐭𝐡𝐞 𝐦𝐨𝐭𝐨𝐫 × 𝐩𝐮𝐥𝐥𝐞𝐲 𝐫𝐚𝐭𝐢𝐨 𝐦𝐢𝐧 𝐍𝟏 = 𝟏𝟒𝟐𝟎𝐫𝐞𝐯/𝐦𝐢𝐧 × 𝟏⁄𝟑 , 𝐍𝟏 = 𝟒𝟕𝟑. 𝟑𝟑 𝐍𝟏 = 𝟏𝟒𝟐𝟎

𝑵𝟐 =

𝟒𝟕𝟑.𝟑𝟑×𝟓𝟎 𝟏𝟔𝟎

,

𝐍𝟐 = 𝟏𝟒𝟕. 𝟗𝟏𝟓 𝐫𝐩𝐦


17


𝐕𝐑 =

𝑵𝟐 𝑵𝟐′

𝐍𝟐′ =

𝟏𝟒𝟕.𝟗𝟏𝟓 𝟏𝟔𝟎

𝐕𝐑 =

𝑵𝟑 =

𝑵𝟐′ 𝑵𝟑

×

×

𝒅𝟐′

(7)

𝒅𝟐

× 𝟕𝟎 , 𝐍𝟐′ = 𝟔𝟒. 𝟕𝟏 𝐫𝐩𝐦

𝒅𝟑 𝒅𝟐′

𝟔𝟒.𝟕𝟏 𝟏𝟔𝟎

(8)

× 𝟕𝟎 , 𝐍𝟑 = 𝟐𝟖. 𝟑𝟏 𝐫𝐩𝐦 .

N3 is the speed required for expressing oil from melon seed. 2.2.3 Selection of Belt Drive The selection of the belt drive was based on the following factors: Speed of the driving and driven shafts, speed reduction ratio, power to be transmitted, centre distance between the shafts, positive drive requirements, shafts layout, space available and service conditions. The length of the belt was calculated using equation (9) (Khurmi and Gupta, 2005). 𝑳 = 𝝅(𝒓𝟏 + 𝒓 𝟐 ) + 𝟐𝒄 +

(𝒓𝟏 +𝒓𝟐 )𝟐 𝑪

(9)

Shaft pulley diameter d2 = 143 mm and so r1=71.5 mm Motor pulley diameter, 𝐃𝟐 = 𝟒𝟖 𝐦𝐦 and so r2 = 24 mm Centre Distance, 𝐜 = 𝟒𝟎𝟎 𝐦𝐦 L = 1123 mm Therefore 1123 mm is selected as the length of the belt. 2.2.4 Shaft Design To design for the shaft one must know the weight of the pulley and gears that will be acting on the shaft as shown in Figure 4.


18


W1

W2

R1

R2 100 mm

50mm

215 mm

Fig. 4. Loading of the Shaft Where: W1 = weight of gear on the shaft; W2 = weight of pulley on the shaft; R1 and R2 = reactions The weight of the pulley is given as W1 = mg

(10)

Where 𝐦 = 𝐦𝐚𝐬𝐬 = 𝐯𝐨𝐥𝐮𝐦𝐞 × 𝐝𝐞𝐧𝐬𝐢𝐭𝐲 g = acceleration due to gravity Density of mild steel is given as 7850 kg/m3 (Khurmi and Gupta, 2005) Volume surface area thickness of the pulley (t) 𝑽=

𝝅𝑫𝟐 𝟒

×𝒕

(11)

Where: Diameter of the pulley on the shaft 𝑫𝟏 = 𝟏𝟒𝟑 𝒎𝒎 = 𝟎. 𝟏𝟒𝟑 𝒎 Thickness (t) = 𝟑𝟎 𝒎𝒎 = 𝟎. 𝟎𝟑 𝒎 𝑾𝟐 = 𝑫𝑽𝒈

(12)

W2 = 7850 kg/m3 x 0.000482 m3 x 9.81 m/s2 = 37.1181 N But weight of gear is measured to be 30 N W1 = 30 N Therefore Fig. 4 is now loaded as shown in Figure 5


19


37.12 N

𝑅2

30 N

65 mm

100 mm mm

50

Fig.5. Loading of the Shaft with the Forces known Now find R1 and R2. ∑ 𝑭𝒚 = 𝟎 𝑹𝟏 + 𝑹𝟐 = 𝟑𝟎 + 𝟑𝟕. 𝟏𝟏𝟖 𝑵 Taking moment about point R2 ∑𝐌 = 𝟎 𝑹𝟏 = 𝟔𝟎. 𝟏𝟖𝟒 𝑵 and R2 = 6.934 N To get maximum bending moment and maximum shear force, we draw the bending moment and the shear force diagrams using the force analysis shown in Figure 6. 30N

37.12N

X1

6.934N

65mm

X2

X1

X2

60.184N

100mm

50mm

Fig. 6: Force Analysis for the Bending Moment and Shear Force Diagrams When 𝐱 = 𝟎 𝐌𝐨𝐦𝐞𝐧𝐭 𝐌 = 𝟎 𝐖𝐡𝐞𝐧 𝐱 = 𝟓𝟎 (𝐟𝐫𝐨𝐦 𝐫𝐢𝐠𝐡𝐭) ∑𝐌 = 𝟎 ∑𝐌 = 𝟑𝟕. 𝟏𝟐 × 𝟓𝟎 = 𝟏𝟖𝟓𝟔 𝐰𝐡𝐞𝐧 𝐱 → 𝟏𝟓𝟎 ↓ ∑𝐌 = 𝟎 ∑𝐌 = 𝟑𝟕. 𝟏𝟐 × 𝟏𝟓𝟎 − 𝟔𝟎. 𝟏𝟖𝟒 × 𝟏𝟎𝟎 = −𝟒𝟓𝟎. 𝟒 𝐍𝐦 𝐰𝐡𝐞𝐧 𝐱 → 𝟐𝟏𝟓 Nigerian Institution of Agricultural Engineers © www.niae.net

20


↑ ∑𝐌 = 𝟎 𝟑𝟕. 𝟏𝟏𝟖 × 𝟐𝟏𝟓 – (𝟔𝟎. 𝟏𝟖𝟒 × 𝟏𝟔𝟓) + (𝟑𝟎 × 𝟔𝟓) = 𝟗𝟗𝟑𝟎. 𝟑𝟕 − 𝟗𝟗𝟑𝟎. 𝟑𝟕 = 𝟎 The Bending Moment and Shear Force Diagrams are shown in Figure 7. 30N

37.118N

6.934N

60.118N 60.118N

37.12N

30N

6.934N

450.7m

Fig 7. Bending moment and shear force diagrams 2.3

Performance Tests

The bara type of shelled egusi melon was used for the performance tests of the machine. The moisture content of the melon pods was determined by gravimetric method and was found to be 6.5 % (wet basis). Tests conducted by feeding 5 kg of melon seeds into the machine, and operating the machine. After each run, materials were collected from the discharge outlet and weighed accordingly. The processing parameters below were determined and used for the calculation of the expeller performance parameter: W1 = Weight of the melon seeds; W2 = Weight of expelled oil; W3 = Weight of cake; T = Time of expelling; Tp = Throughput; Er = Extraction rate; Y = Yield; Ee = Extraction efficiency. Throughput 𝑇𝑝 (𝑘𝑔/ℎ) = Weight of melon seeds fed into expeller ......................(13) Total time taken to extract oil from the melon seeds


21


Extraction Rate Re (Lt/h) /(𝑘𝑔/ℎ) = Weight/ litre oil extracted......................(14) Time of extraction Yield Ye (%)

=

litre oil obtained x 100..………………………………... (15) Original weight of seeds

Extraction efficiency Ee (%) = litre of oil obtained x 100............................ (16) Oil content of the melon seeds 3.

RESULTS AND DISCUSSION

From Tables 1 – 3, the oil yield of 30.23%, 1.52 litre expression and efficiency of 60.83% were obtained for the motorized expeller while that of the traditional method gave oil yield of 16.70L /kg and efficiency of 33.43%. On the average a total time of 30.18 minutes was required by the motorized expeller to expel 1.52 litres of oil from 5 kg of melon seeds while the traditional method used 2 h 38min to expel an average of 0.84 litres of oil from 5 kg of melon seeds. It can be seen that the motorized melon seeds oil expeller performed far better than the traditional method. Further analysis reveals that these differences in the averages of the selected variables of the two groups are statistically significant at 1% level (Table 4.0). This suggests that the litre of oil gotten from motorized method for example is significantly higher than that gotten from traditional method. This is also true for the yield. The same conclusion is drawn for throughput and efficiency. On the other hand, it takes longer time for the expelling process in traditional method than motorized method. Table 1. Performance Evaluation Indices of Motorized Melon Oil Expeller at Constant Motor Speed (1420rpm), Moisture Content (6.0%) and Constant Weight of Melon Seeds (5kg) Time of Ext. Litre of Oil Wt of Cake Throughput Yield Efficiency T (min) (Lt) W2 (Kg) Tp (Kg/h) Ye (%) Ef (%) 30.00 1.50 3.20 10.00 30.00 60.00 29.50 1.54 3.18 10.16 30.80 62.00 30.00 1.55 3.20 10.00 30.40 62.00 29.50 1.51 3.20 10.20 30.20 61.00 29.55 1.50 3.30 10.15 30.00 60.00 29.50 1.50 3.25 10.20 30.00 60.00 30.18 1.52 3.22 10.12 30.23 60.83 Table 2. Performance Evaluation Indices of Traditional Method of Melon Oil Expression Time of Ext. Litre of Oil Wt of Cake Throughput Yield Efficiency T (h) (L) W2 (Kg) Tp (Kg/h) Ye (%) Ef (%) 1.30 1.10 4.20 3.33 22.00 44.00 1.50 0.90 4.15 2.73 18.00 36.20 2.00 0.86 4.30 2.50 17.20 34.40 2.20 0.80 4.35 2.15 16.00 32.00 2.35 0.75 4.45 1.94 15.00 30.00 2.55 0.60 5.25 1.71 12.00 24.00 Mean 1.98 0.84 4.45 2.39 16.70 33.43 Nigerian Institution of Agricultural Engineers © www.niae.net

22


Table 3. Summary of Statistical Analysis on Motorized and Traditional Methods of Melon Oil Extraction Parameter

Type of Method

Litres of oil

Weight Throughput Time Yield Efficiency

No of times

Mean

Std. Mean Error

Motorized method Traditional method Motorized method Traditional method Motorized method Traditional method Motorized method

6 6 6 6 6 6 6

1.5167 0.8350 3.2217 4.4500 10.0833 2.3933 29.74

0.00919 0.06801 0.01833 0.16583 0.02940 0.24040 0.09167

Traditional method Motorized method Traditional method Motorized method Traditional method

6 6 6 6 6

131.67 30.2333 16.7000 60.8333 33.4333

12.6920 0.13081 1.36015 0.40139 2.72686

Table 4. Two-tail Comparison Test for the Methods of Melon Oil Extraction Variable Litres of oil Weight Time Throughput Yield Efficiency *significant at 1%

T 9.933 7.362 8.030 31.752 9.904 9.941

Df 10 10 10 10 10 10

Sig. (2-tailed)* 0.001 0.001 0.001 0.001 0.001 0.001

Mean Difference 0.68167 -1.22833 -101.950 7.69000 13.53333 27.40000

Std. Error Difference 0.06863 0.16684 12.693 0.24219 1.36642 2.75625

Table 4 shows that there is a significant difference at 1% level between the performance indices of the developed melon oil expeller and those of the traditional method of melon oil expression. The difference between the traditional and motorized method with respect to oil yield, efficiency and time of expression are shown in fig 1 – 3. The figures show that the motorized method is better than the traditional using the parameters (oil yield, efficiency and time required for expression.


23


1.80 1.60

Litre of Oil (L)

1.40 1.20 1.00 0.80

Motorized

0.60

Traditional

0.40 0.20 0.00

1

2

3

4

5

6

Fig.1. Effect of Liters of Oil Expression on number of Replications from Melon Seeds using Motorized and Traditional Methods

70.00

Efficiency (%)

60.00 50.00 40.00

Motorized

30.00

Traditional

20.00 10.00 0.00

1

2

3

4

5

6

Fig .2. Effect of Efficiencies on number of Replications on Motorized and Traditional Methods of Expression oil from Melon Seeds


24

Time of Extraction (Min)


180.00 160.00 140.00 120.00 100.00 80.00 60.00 40.00 20.00 0.00

Motorized Traditional

1

2

3

4

5

6

Fig.3. Effect of Time of Oil Expression on number of Replications by using Motorized and Traditional Methods 4.

CONCLUSIONS

A motorized melon oil expeller was designed, fabricated and performance evaluation test carried out. The main essence of the machine was to provide a means that would enhance the processing of melon seeds to obtain oil thereby reducing the drudgery and time-consuming operations when the traditional method is used. The result showed that the machine attained oil expelling of 1.52litres, oil yield of 30.23% and efficiency of 60.83% when 5 kg of melon seeds was used as basis. REFERENCES Adeniran, M.O. and Wilson, G.F. 1981. Seed Type Classification of Egusi Melon in Nigeria. Paper presented at the 6th African Horticultural Symposium, University of Ibadan, 9th –25th July, 1981. Douglas, M. C. and Glenn, D. 1982. Foods and Food Production Encyclopedia. Van Reinho. D, New York. Khurmi, R. S. and Gupta, J. K. 2005. A Textbook of Machine Design. Eurasia Publishing House, New Delhi. NHORT (National Horticultural Research Institute) 1976-1986: Advances in Fruits and Vegetables Research at NIHORT; A Commemorative Publication to mark the 10th Anniversary of the NIHORT Ibadan Odigboh, E.A. 1979. Impact Egusi Shelling Machine Transaction of America Society of Agricultural Engineers 22(5):1264-1267. Omidiji, M. O, et al 1985. Exploratory Survey on Cropping Systems and Related Activities at Ilugun Local Government Area, Ogun State. In: Farming Systems Research in Nigeria; Diagnostic Surveys. Omidiji, M. O. 1997. Tropical Cucurbitaceous Oil Plants of Nigeria. Vegetables for the Humid Tropics. A Newsletter and Annual Communication among Research Workers. No 2: 37 –39.


25


DEVELOPMENT OF A TOMATO SLICING MACHINE E. C. Oriaku, C. N. Agulana and N. I. Nwagugu Engineering Research and Production Department (ERDP) Projects Development Institute (PRODA), Enugu E-mail: [email protected] ABSTRACT In many parts of Nigeria, post-harvest and storage spoilage losses are enormous because of poor handling facilities for agricultural produce. This problem leads to undue scarcity of agricultural produce immediately after harvesting season. In order to address this problem, a tomato slicing machine was designed, fabricated and tested. This machine is intended to add value to ripped tomatoes by slicing and exposing the water cells for easy drying in a dryer on an industrial scale. The sliced dried tomato chips gotten will be packaged for storage in order to improve its shelf life. The performance evaluation of the fabricated machine was carried out to be very efficient. Several quantity of measured tomatoes were fed into the machine in a given time taking about 2kg in a minute under a continuous feeding. The feed and slicing time were observed and measured. The results of the tests were tabulated and analyzed. The machine provides a lasting solution to the drudgery associated with tomato slicing. KEYWORDS: Design, fabrication, testing, tomato slicing, processing. 1.

INTRODUCTION

Tomato (Solanum lycopersicum) is one of the most important vegetable plants in the world. A major vegetable crop that has achieved tremendous popularity over the last century. It is grown in practically every country of the world - in outdoor fields, greenhouses and net houses. It is said to have originated in western South America, and domestication is thought to have occurred in Central America. (Zvi H. W 2000). Because of its importance as food, tomato has been breed to improve productivity, fruit quality, and resistance to biotic and a biotic stresses. Tomato has been widely used not only as food, but also as research material. Tomato is a major agricultural crop cultivated in Nigeria especially in the Northernpart. Oyeniran (1988) reported that in Nigeria, figures like 6 million tonnes of tomatoes have been given as annual production level. During harvesting period, several tones of this tomato fruit are produced. Unfortunately, they are not only seasonal but highly perishable and deteriorate very fast loosing almost all the required attributes and some may likely result to total waste because they cannot be stored for longer duration. It has been shown that as high as 50% of these produce are lost between rural production and town consumption in the tropical areas (Oyeniran, 1988). Due to perishability, farmers are losing a bulk of produce each year. Although much is consumed locally and a little exported to other neighbouring African countries, yet large wastages is still incurred during each harvesting season. This reason coupled with the fact that tomato is a seasonal produce, causes the scarcity of this fruit to step in immediately the harvesting period is over. Tomato is a very nutritive fruit with moisture content of about 90% (Akpinar et al., 2003). Tomatoes, aside from being tasty, are very healthy as they are a good source of vitamins A and C. It also helps maintain capillaries, bones and teeth and aids in the absorption of iron. It is grounded and used for several food preparations. It can be eaten raw. Also, it can be sliced and be used for salad and confectioneries. Tomato could be sliced and dried for storage purposes to take care of the non-harvesting seasons and also to reduce wastage. In developing countries the tomato is becoming a more important part of the food basket but the goal of the farmer is to produce quantity not quality so that people can eat. (Zvi H. W 2000). It is estimated that 45 million tonnes of tomatoes are produced each year from 2.2million hectares excluding the large amount grown in home gardens.(Villareal,1980)


26


The essence of the slicing before drying is to expose the water cells of the tomato fruit for easy drying. After drying and cooling, the tomato could be packaged for storage. Furthermore, it could be milled to dust and packaged as grounded dried tomato for easy cooking. The objective of this project work is therefore, to design, produce and test a tomato slicing machine as one of the equipment that will facilitate the drying and packaging of dry tomato in order to reduce wastage incurred during harvesting period and also make the tomato fruit last longer during storage in order to take care of the scarcity periods of this particular fruit. 2.

MATERIALS AND METHOD

2.1

Description of the Tomato Slicing Machine

The tomato slicing machine is shown in Fig. 1. The main components of the machine: Sleeve with knives; Slicing chamber; Discharge chamber; Feed hopper; Structural frame; Slicing bed; Discharge spout; Transmission shaft; Knives carrier; Ball bearings; Pulleys; Belts; Dividing strips; Electric motor; Main casing with reinforcement. Sharp strips of flat plates are arranged on long sleeve in opposite direction as knives. The sleeve is bolted at the centre of the transmission shaft. The shaft is mounted at the inside centre of the slicing chamber via bearings at the outside of the chamber. The exterior of the shaft on one side is connected to the driven pulley. The driver and the driven pulleys are then connected with a v-belt. The slicing bed acts as a support for the tomatoes for easy slicing. The slicing bed is a rectangular formation with bolting points at the four edges. The inside of the bed is divided with strips of metals giving gaps for the rotation of the knives in between the gaps. The division is according to the number of knives. The gap between a knife and the bed strip is reasonable enough to allow for the passage of sliced tomatoes. The inlet hopper is arranged to direct and spread the incoming unsliced tomatoes to the slicing beds. One tomato can be split into about four pieces depending on the size. 2.2

Mechanics of Operation of the Tomato Slicing Machine

All the components of the machine are systematically arranged on the structural stand. The mechanics of operation of this machine is purely based on the dynamics of the components namely pulleys, belt, shaft, bearings etc. The electric motor provides the primary motion which is transmitted via pulleys, bearings and belt to the shaft and sleeve carrying the knives. The rotational motion of these components, gravitational motion of the tomato fruit through the hopper outlet channel are employed in order to achieve the desired slicing operation. The knives rotate at a medium speed of (1440) rpm in between the dividing stripes on the bed. Slicing of the tomatoes is achieved on the bed with this speed. The rotary knives are carried on a sleeve which is bolted on a shaft. There are thirty two (32) number of knives. The dimension of one knife is shown in Fig. 2. The thickness of the knife is 2mm. The edges are sharpened to aid slicing.

L W Fig 2. Knife Dimensions


27


Fig. 1. The Tomato Slicing Machine 2.2.1 Rotational Motion and Centrifugal Force (Fc) The rotational motion from the shaft of the prime mover (electric motor shaft) is transmitted to the input shaft carrying the driven pulley and knives. For any object of mass M moving in an circular motion, its acceleration is directed towards the centre of the body and its linear velocity is tangential to the radius of the object. The displacement which starts from point A, then to B and continues is in terms of θ. The angular velocity is designated ω. The acceleration (a) of the rotary body is given as (Fig. 3): a=ω2r………………………………………………….. (1)


28


(J. Hannah & R. Stephens, 1984) Fig 3. Centrifugal Force Where r = radius of the object. The acceleration is centripetal. The radially inward, or centripetal force required to produce acceleration is given as Fc

=

Ma = Mω2r = MV2/r ………………………………………...(2)

(John Hannah &R.C. Stephens 1984) If the body rotates at the end of an arm, this force is provided by the tension on the arm, the reaction of this force acts at the centre of rotation and is centrifugal force. It represents the inertia of the body resisting the change in the direction of the motion. A common concept of centrifugal force in engineering problems is to regard it as radially outward force which must be applied to a body to convert the dynamical condition to the equivalent static condition. 2.2.2 Rotational Torque (T) The value of torque developed by a rotational body is given as the product of the force causing the motion multiplied by the radius of rotation (John Hannah &R.C. Stephens 1984) T = FC x r ………………………………………………………………………….(3) Where: FC = Centrifugal Force, R = radius 2.2.3 Work Done by a Torque If a constant torque T moves through an angle θ Work done = T ‫ ٭‬θ ………………………………………………………………. (4) If torque varies linearly from zero to a maximum value T Work done = ½ T ‫ ٭‬θ ………………………………………………………………(5) In general case where T = f (θ) Work done = ʃ f(θ)de …………………………………………………………… (6) The power (p) developed by a torque T(N.M) moving at ω rad/sec is P =Tω = 2π(watts) ………………………………………………………………. (7) Where N is the speed in rev/min and ω=

2𝜋𝑁 𝜖𝑜


29


2.2.4 Pulley and belt drive The schematic representation of the pulley system is shown in Fig. 4.

θ

θ

Fig 4. The Pulley System The velocity ratio between two pulleys transmitting torque or power is given as 𝑊1 𝑊2

𝑁

𝐷

= 𝑁1 = 𝐷12 2

𝑁

i.e. D2 = ( 𝑁1 ) D1 = n D1 ………………………………(11) (Avallone E.and Baumeister T.,1997) 2

where: W1 = angular velocity of driver; W2 = angular velocity of driven; N1= rpm of driver; N2 = rpm of driven D1 = diameter of the driver; D2 = diameter of the driven; θ = angle of lap between the belt and the pulley; n = N1/N2 = speed ratio 2.2.5 Tension on Belt Driver (T1 and T2) For belt transmission between two pulleys (Hall et al.,1961) 𝑇1 𝑇2

= е𝜇𝜃 ……………………………………………………….…………..(8)

Also 𝑇1 −𝑇𝑐 𝑇2 −𝑇𝑐

=е𝑚𝜃 ………………………………………………………………....(9)

And Tc = mv2 ………………………………………………………..……..(10) 2.2.6 Power Requirement (P) of the Slicing Machine The power requirement of the slicing machine is given as: P = (T1- T2) v…………………………………………………………………..(11) In this equation, the power (P) is in Watts, when T1 and T2 are in Newton and belt velocity is in meter per second. When the tensions are in pounds and the velocity in feet per minutes, the horse power(Hp) transmitted is given as Hp =

(𝑇1− 𝑇2)𝑣 33000

…………………………………………………………..…..(12) (Hall et al.,1961)

Where: T1 = tension on tight side of belt; T2 = tension on slack side of belt; Tc = centrifugal tension; W1= angular velocity of driver; W2 = angular velocity of driven; θ = angle of lap between belt and pulley; v = linear velocity of belt


30


2.2.7 Belt Length (L) This is given as L=

𝜋(𝐷1+𝐷2 )

(𝐷1−𝐷2

2

4𝑐

+

+ (2𝑐) …………………………..(13) (Avallone E. and Baumeister T.,1997)

C = centre distance of two pulleys 2.2.8 Shaft Design The shaft with the forces acting on it is represented schematically in Fig. 5. centrifugal force

Load due to weight of tomato

Bearing reaction (1)

Bearing reaction (2)

Fig 5. Force analysis From the evaluation of the forces and determination of bearing reactions, the maximum bending moment of the shaft is evaluated (Mmax). The shaft diameter (D) is calculated using ASME code standard for shafting. The standard equation for shafting is stated below 5.2 𝜏𝑑

D= { [(Cm x Mmax)2 + (CT x T)2]1/2}1/3 …………………………………….(14) For ASME code standard 𝝉𝒅=𝟎.𝟑𝜹𝒚 𝒐𝒓 𝟎.𝟏𝟖𝜹𝒖 ………………………………………………….(15) The smaller of the two is chosen as 𝝉𝒅 . The presence of key sit on the shaft reduces the value of 𝝉𝒅 by 75%. For rotating shafts Cm 1.5, Cr = 1 Where: D = diameter of shaft; 𝝉𝒅 = allowable shear stress; Cm = moment factor; CT = Torque factor; Mmax = maximum bending moment; T = rotational torque; 𝛿𝑦 = yield stress of shaft material; 𝛿𝑢 = ultimate stress of shaft material. Material used for shafting is stainless steel 2.2.9 Bearing Selection (Loading) Bearing selection was carried out using the following equations by Shigley and Mtscake (1961) Pb = XVFr + YFa ………………………………………………………………………….(16) L10 = [C/Pb ]3 ……………………………………..……………………………………...(17) Nigerian Institution of Agricultural Engineers © www.niae.net

31


Where: Pb = bearing load; Fr = radial load; Fa = axial load; X = radial load factor; C = basic load rating; V = Inner ring rotation factor; L10 = bearing life in million revolution. 2.2.10 Slicing Bed The designed tomato slicing machine has a rectangular slicing bed with metal partitions. The knives rotate in these partition. The tomatoes to be sliced fall from the hopper of the machine by gravity to this bed. The bed supports the tomatoes and makes for easy slicing. The sketch of the bed is shown in Fig.6.

Fig 6. The Slicing Bed 2.3

Testing of the Slicing Machine

2.3.1 Free Test Run of the Machine The machine was tested in the Department of engineering research and development projects, (ERDP) Fabrication Workshop PRODA, ENUGU. The test was carried out in two different stages. Stage (1), the free test run (without load) and stage (2), testing with varied loads (i.e fresh tomatoes) with different weights of (5.0kg, 8.0 kg, 12kg, 15kg to 58kgrams) respectively. A stop watch and a weighing balance were used to ascertain the feed time in seconds, duration of slicing and measuring the weight of tomatoes to be sliced, receptively. Analyses of the results were made. 2.3.2 Actual Test on Load The tomato that was used for the test running of the Slicer was gotten from an open market at Ogbete main market Enugu state of Nigeria. Also, some tomato fruit of long shelf-life hybrid, were obtained from a commercial farm in the area of Obinagu located less than 2km south of the Institute. Fruits were harvested early in the morning and transported to the laboratory for the experimentation. Fruits were sorted and washed. There was no treatment given to the tomatoes before slicing. Fresh selected tomatoes were used. The samples of the tomatoes used and sliced are shown in Fig. 7.below:


32


Fig 7. Sample of whole and sliced tomatoes 2.4

Performance Evaluation of the Tomato Slicing Machine

The tomato slicing machine was tested by feeding several quantity of measured tomato into the machine in a given time. The feed and slicing period (time) were observed, measured and tabulated. 3.


The results of the performance evaluation are shown in Table 1 and Figs 8 – 11. Table 1: Performance Evaluation of Tomato Slicing Machine. S/N Weight of Feed time Feed rate Slicing time tomato (kg) (s) (kg/s) (s) 1 5 3 1.67 4

Slicing rate (kg/s) 1.25

2

8

4

2.00

5

1.60

1

3

12

5

2.40

7

1.71

2

4

15

6

2.50

9

1.67

3

5

19

8

2.38

12

1.58

4

6

26

12

2.17

17

1.53

5

7

30

15

2.00

20

1.50

5

8

40

22

1.82

28

1.43

6

9

50

28

1.79

36

1.39

8

10

58

34

1.71

44

1.32

10

137 13.7

20.44 2.044

182 18.2

14.98 1.498

45 4.5

Total Mean

263 26.3


Difference between feed and slicing time (s) 1

33


40

weight of tomato(kg)

35

34

30

28

25 22

20 15

15 12

10 5

5

4

3

8

6

0 5

8

12

15

19

26

30

40

50

58

Feed time(s)

Fig 8. Effect of weight of tomato on seed time 40

weight of tomato(kg)

35

34

30

28

25 22

20 15

15 12

10 5

4

3

8

6

5

0 5

8

12

15

19

26

30

40

50

58

slicing time(s)

Fig 9. Effect of weight of tomato on slicing time


34


50 45

44

40 36

35 30

28

slicing time(s)

25 20

20 17

15 12

10 5

7

5

4

9

0 3

4

5

6

8

12

15

22

28

34

Feed time(s)

Fig 10. Relationships between slicing time and feed time weight of tomato(kg)

feed time(s)

slicing time(s)

140 120

Time(secs)

100 80 60 40 20 0 1

2

3

4

5

6

7

8

9

10

Quntity (kg)

Fig 11. Relationships among weight of tomato, the feed time and the slicing time From the above graph, it shows that the feed rate calculated from the mean value = 26.3 / 13.7 = 1.91kg/min which is approximately 2kg/min The mean slicing rate per minute is given as 26.3/ 18.2= 1.44kg/min From the table and graph of performance analysis of the tomato slicing machine on different loads, the following inferences can be drawn. Nigerian Institution of Agricultural Engineers © www.niae.net

35


I) There is progressive increase in slicing time as the load increases. II) There is progressive increase in the difference between feeding time and slicing time III) From calculation of the feed rate and slicing rate based on the mean value of load, the slicing rate is close to the feed rate. There is only a 0.47kg difference between the two (2) values. IV) Based the calculated slicing rate, the machine capacity in slicer is put at 2kg per minute. 4.

CONCLUSIONS

The ripped tomatoes after slicing, will no doubt reduce waste of labour and resources if merged with efficient means of drying and packaging to prolong storage. Based on the slicing rate by the machine, high productivity in slicing is achieved and the slicing efficiency is very high and to upward of 90%. The machine is actually useful in minimizing wastage and improving the shelf life of ripped tomatoes. This machine is recommended to all agricultural and food processing industries of the contemporary society since tomato slicing as a unit in tomato preservation and processing operations requires a mechanized means. REFERENCES Avallone E.A.M, and Baumeister T, 1978. Marks standard Handbook for Mechanical Engineers. McGraw Hill Company, New York. Akpinar et al., 2003. Thin layer drying of red pepper. J. Food Eng. v59. 99-104. Aviara,N.A.,Shittu,S.K.,and Haque.M.A. 2007. Physical Properties of guna fruits relevant in bulk handling and mechanical processing Int. Agrophysics, 21,7-16 Edward Shigley J and Charles R.Miscake 1961. Mechanical Engineering Design. McGraw Hill,New York. Hall A.S Jr.,Holowenko A.R.,H.G.Langhin 1961. Machine design schaum’s outline series.McGraw Hill.http://www.soyatech.com/canola-facts.htm. Accessed January 2010 John Hannah &R.C. Stephens 1984. Mechanics of machines. Edward Arnold Oriaku E.C.,Agulana C.N.,and Nwagugu.,N.I. 2010. Research assignment executed in ERDP department of Project Developments Institute (PRODA) Enugu. Oyeniran,J.O. 1988. Reports of the activities of nationally coordinated team on improved packaging and storage of fruits and vegetables in Nigeria. Proceedings of the workshop on improved packaging and storage systems for fruits and vegetables in Nigeria held in Ilorin, Nigeria. Paul H. Black ,and Eugene Adamus Jr. 1982. Machine Design. MacGraw Hill Villareal, R.L. 1980. Tomatoes in the tropics. I.A.D.S. development oriented Literature, series. West View Press, Boulder, Colorado. Zvi Howard Wener 2000. www.agrisupportonline.com. Accessed June, 2011


36


PALM KERNEL OIL EXPRESSION BY UNIAXIAL COMPRESSION I. C. Ozumba and K. Oje National Centre for Agricultural Mechanization, (NCAM), P.M.B 1525, Ilorin, Nigeria. E-mail: [email protected] 2 Agricultural and Biosystems Engineering Department, University of Ilorin, Nigeria.

1

ABSTRACT Palm kernel oil yield was found to be dependent on operative processing variables including: heating temperature, applied pressure and particle size to which ground palm kernel samples of fine and coarse particle sizes were subjected to using a Piston-cylinder Rig under an instrumented hydraulic uniaxial compression. The effects of heating temperature and applied pressure on the oil yield was investigated at heating temperatures of 70, 90, 110, and 130 0C and applied pressures of 6, 9, 12 and 15 MPa. The oil expressed from each temperature level at optimum applied pressure of 15 MPa was subjected to laboratory analysis using AOAC 1991 recommended method at the analytical and quality control laboratory of the Global Soaps and Detergents Industries, Ilorin, Nigeria. From the results, the highest palm kernel oil yield obtained was 32.15 % at optimum conditions of 15 MPa applied pressure for 10 minutes and 110 °C heating temperature for 30 minutes on samples of 4.5 % moisture content dry basis. This yield corresponds to an expression efficiency of 71.8 % of the total palm kernel oil content. Increase in oil yield was recorded at corresponding increase in heating temperature and applied pressure, but the oil yield begins to drop after an optimum heating temperature of 110 °C was attained. Oil yield from fine particle size of palm kernel was discovered to be significantly different from that obtained from coarse particle size at 5 % level of confidence. The result of the physico-chemical characteristics of the palm kernel oil at temperature of between 70 130 0C are: Free fatty acid 4.02 - 4.63 %; Acid value 11.29-12.0 mgKOH/g; rancidity index 2.4R.21Y– 1.9R.10Y; saponification value 246.57-243.80mgKOH/g; viscosity 140 – 60cps and specific gravity 0.908-0.894. The result indicates that the quality of the expressed oil decreases as the heating temperature increases. KEYWORDS: Palm kernel oil, oil yield, expression efficiency, hydraulic uniaxial compression. 1.

INTRODUCTION

The increase in the world’s population has no doubt increased the demand for fats and oils obtained from oil bearing crops. Oil-bearing crops are classified into three, namely: Oil seeds and beans (e.g. Cotton, Rape, Mustard-seed, Beniseed, Sesame, Sunflower, etc); nuts (e.g. Coconut, Copra, groundnut, Shea nut, palm kernel, etc); and mesocarps or fruits (e.g. Oil palm, Avocado, Olive, etc). About 90% of the oil and products from it are used for food applications, while about 10% goes into non food applications, (Pantzaris, 2000). The Oil Palm, which gives both Palm Oil (PO) and Palm Kernel Oil (PKO), is Elaeis guineensis (Hartley, 1988). The Palm Oil (PO), which is reddish in colour, is obtained from the Orange colour mesocarp, while the PKO is obtained from the hard-liquefied cell within the nut, called the kernel. According to Praven, (1997) and Breadson, (1983), modern Processing of Oil bearing crops (seeds or nuts) into edible or industrial oil is practiced using different methods, which may be categorized into three. One is the solvent extraction method in which a solvent, when brought in contact with the preconditioned oil seed or nut, dissolves the oil present in the oil bearing material and the separated mixture is later heated to evaporate the solvent and obtain the oil. Mechanical oil expression is the second method. In this process, the preconditioned oil seed or nut is passed through a screw press, a hydraulic press or a ram press, where a combination of high temperature and pressure is used to crush the Nigerian Institution of Agricultural Engineers © www.niae.net

37


oil bearing material to release the oil. The third method is the wet processing in which the oil bearing material is boiled in water leading to a partial separation of oil (clarification). The effect of particle size, heating temperature, heating time, applied pressure and duration of pressing on the yield and quality of mechanically expressed groundnut oil were investigated by Adeeko and Ajibola (1989). Result showed that oil yields from coarsely ground groundnut samples were higher than those from finely ground samples but the free fatty acid values were lower. Increasing the temperature did not improve the oil yield and the rate of oil expression was increased by an increase in temperature, time of heating, and particle size. An increase in the heating temperature and heating time increased the free fatty acid value, peroxide value and the colour intensity of the oil expressed. Olaniyan (2010) investigated the effects of some process conditions like nature of bean, heating temperature and pressing time on the yield and quality of oil mechanically expressed from castor bean using a piston-cylinder rig in association with California Bearing Ratio Universal Testing Machine (CBR – UTM). The Results showed that process conditions and their interactions were significant on oil yield at 0.05 % level of significance. However, only the pressing time was significant on the extraction pressure. Oil yield increased with increased heating temperature and pressing time for the crushed bean, shelled bean and unshelled bean. Bamigboye and Adejumo (2011) carried out investigation on the effects of the processing parameters of Roselle Seed on its oil yield. The seeds were ground and classified into two particle sizes (fine and coarse). The samples were conditioned by adding calculated amount of distilled water to different moisture levels from the initial moisture levels of 4.4 % and 5.14 % respectively. The samples were heated at different temperatures of 80, 90,100 and 110 over a period of 15, 20, 25 and 30 minutes, expressed at 15, 22.5, 30 and 37.5 MPa using hydraulic oil extractor for 10, 20, 30 and 40 minutes. The result of the investigations showed that oil yield increases by 5 % - 6 % with an increase in the processing parameters of pressure up to 30 MPa, temperature of 100 and decreased beyond these points. Finally ground samples were found to have a higher yield than coarsely ground samples at the different processing parameters. They concluded by affirming that processing parameters affect the oil yield from Roselle Seeds. Not much can be found in the literature on studies undertaken on processing factors as related to oil yield from ground palm kernel of Dura variety. This prompted the present work to study the effects of some processing parameters on palm kernel oil expression. 2.

EXPERIMENTAL PROCEDURE

All experimental investigations were carried in the Engineering Materials Testing laboratory of the Engineering and Scientific Services (ESS) department of the National Centre for Agricultural Mechanization (NCAM), Ilorin, Nigeria. The room temperature of the laboratory throughout the duration of the experimental works was averagely 30 0C. 2.1

Material Preparation

Palm Kernel (Dura Variety) used in this experiment was obtained from the Nigerian Institute for Oil Palm Research (NIFOR), Benin city, Nigeria. Moisture content of the palm kernel at the point of procurement was determined and found to be 11.5 %. The kernels were further dried to 4.5% moisture content using sun drying method. The kernels were cleaned to remove stones and other foreign materials; after which they were firstly crushed using a hammer mill, and later the crushed meal was further reduced using an attrition mill as reported by Olaniyan, (2006). They were later packed into air tight containers and stored in the laboratory.


38


2.2

Experimental Machines and Instrumentation

Oil expression process normally involves the application of compressive force on the oil seed flakes enclosed in a suitable retaining envelope. 14

1 14 13 12 11 10 9 8 7 6 5 4 3 2 1

2 3

13

4 5 12

Stod Heater band Thermocouple Base plate Load cell with data cable Oil collector Supporting plat form 8mm bolt Threaded ring Press cage cylinder Heater band plug port Plug (from AC orDC means) Heater band clip Compression piston

6 7 8

9

11

10

Fig. 1: Exploded view of the Model laboratory Mechanical Oil Expression Rig Laboratory mechanical oil expressing piston-cylinder rig was modified and fabricated, and used for this investigation (see Fig. 1). The mechanical oil expression rig, which is similar to the one used by Olaniyan (2006) Mrema 1979 (and reported by Mrema and McNutty 1980, 1984 and 1985 on mechanical expression of oil from rape seed, cashew and Shea butter) is made up of three major components: the compression piston, the press cage cylinder and the supporting platform (Olaniyan 2006). A 600 W electric band heater was installed round the press cage cylinder to serve as a heating device for the expression process. The rig was adequately instrumented with a temperature controller to control the expression temperature, while the pressure for oil expression was obtained from the hydraulic press (made in Denmark Stenhous A/S hydraulic press with model number; 52773703 of 15 tons capacity and actuated using a lever) via the instrumentation system. The temperature controlling exercise was achieved with thermocouple connected to an Electronic Temperature Controller (Model JTC-902), which was designed and manufactured in Japan. Heat sensing was achieved by inserting the thermocouple probe into the oil seed sample through a hole on the press cylinder. A complete assembly of the Hydraulic press, the mechanical oil expression rig and the instrumentation system used throughout for the experimental investigation is as shown in Figure 2.


39


2.3

Experimental Investigation Procedure

2.3.1 Analytical Procedure The oil content of ground palm kernel was determined using Soxhlet extraction apparatus in the quality control laboratory of Global Soap and Detergent, Ilorin, Nigeria. This is in accordance with the Association of Analytical Chemists (AOAC) direct gravimetric method of Soxhlet extraction.

A

B

C D

Fig. 2: Complete Assembly of the Force Measuring Device, the Mechanical Oil Expression Rig with the Temperature Controller on Hydraulic Press. Legend A Mechanical Oil Expression Rig B Temperature Controller C Amplifier with display Unit( Force Measuring Device) D Load Cell Likewise the physic-chemical characteristics of the expressed oil at the various heating temperature was determined using the American Oil Chemists Society (AOCS) method. Particle size analysis was carried out using a set of laboratory Endocotts Test Sieves and Shaker (model SW19 3BR England) available in the Laboratory of the Soil and Water Management Engineering Department of NCAM. Particle size analysis was done in accordance with ASAE (1989) Standard S319. In line with Adeeko and Ajibola (1990), samples that passed through the 5.6mm sieve but retained on the 2.36mm sieve were classified as coarse, while samples that passed through 2.36mm sieve but retained on the 0.6mm sieve were classified as fine. Moisture content of the sample was determined by oven drying 100g of ground sample at 130 0C for 6 hours; as recommended for oil seeds by Young et al (1982) and used by Tunde-Akintunde et al (2001).


40


2.3.2 Heating Sample Heating of the sample (milled palm kernel) was achieved by weighing 200g of the sample in line with Olaniyan (2006), and transfer of the weighed sample into the press cage already encircled with the temperature controlled heater band (see Fig. 3). The samples in the press cage where heated to temperatures of 70 0C, 90 0C, 1100C and 130 0C respectively for 30 minutes before expression begins. The lower limit of 70 0C and upper limits of 130 0C heating temperature was selected based on preliminary laboratory investigations, which revealed that heating milled palm kernel sample below 70 0C did not give good oil yield during expression; while heating above 130 0C results in excessive burning and darkening of the oil.

A B C D

Fig. 3: Assembly of the heating system Key: A= press cage cylinder, B=heater band, C=thermocouple,

D=temperature controller

Also, the heating time of 30 minutes used in this study was chosen based on preliminary investigations and also on the fact that the period allows for temperature uniformity and equilibration of the oil seed cake as reported by Hamzat and Clarke (1993). 2.3.3 Sequence of Mechanical Oil Expression The complete assembly of the hydraulic press, the mechanical oil expression rig with the temperature regulator, and the compressive force measuring device used in this experiment is as shown in figure 2.3. Before coupling the mechanical oil expression rig, a stainless steel wire mesh was placed at the bottom of the cylinder guide in order to cover the drainage area and at the same time serve as a filter during the oil expression process. After the coupling, a sample of 200g weight of milled palm kernel was poured into the press cage cylinder. The sample was then heated for 30 minutes at heating temperature of 700C. Using the actuating lever of the hydraulic press, the compression piston was moved down to touch the sample and precompact it to a height of 70 mm (Olaniyan, 2006) inside the press cage cylinder. After the precompaction, the sample was further compressed by the hydraulic press via the compression piston at pressures of 6.0 MPa, for 10 minutes. The oil expressed drains into the oil collector and was collected through the outlet pipe. On the expiration of the pressing time, the compression piston was lifted above the press cage using the actuating lever of the hydraulic press. The press cage cylinder was then unscrewed and the residual cake was extruded and weighed using a precision electronic weighing balance 1kg (0.005g) BC-Series, Manufactured by ORMA, Germany. The same procedure was followed to carry out the experiment


41


for three other heating temperature levels of 90 0C, 110 0C, and 130 0C at 9.0, 12.0, and 15.0 MPa respectively. 2.3.4 Determination of Physical and Chemical Characteristics of Expressed Oil at Different Temperatures The physical and chemical properties of the expressed oil determine the quality of the oil. The physical and chemical analyses of each sample of oil expressed at heating temperatures of 70 °C, 90 °C, 110 °C, and 130 °C were carried out using the AOAC (1991) recommended method at the Analytical and Quality Control Laboratory of the Global Soaps and Detergents Industries Limited, Asa Dam Road, Ilorin, Nigeria. The physical and chemical properties of the expressed oil determined are: Free fatty acid, Acid value, rancidity index, Saponification value, viscosity and specific ravity. 3.


3.1

Effects of Heating Temperature on Oil Yield

Table 1 is the summary of the result obtained for oil yield, at the various process conditions, while table 2 is the summary of ANOVA showing the effect of the process conditions on oil yield during mechanical expression of oil from palm kernel under uniaxial loading. From the ANOVA table, it can be observed that the process conditions (that is heating temperature and applied pressure) and their interactions had significant effect on oil yield, at 1% level of significance for the fine and coarse particle sizes respectively. Hence, the hypothesis of equality of mean treatment effect is rejected, and it can be implied that at least one of the mean treatment effect is significantly different from the others. In order to determine the differences in the mean treatment effect of heating temperature on oil yield for both particle sizes, New Duncan’s Multiple Range Test (NDMRT) was carried out. The result of the comparison among the four levels of heating temperature for each of the particle size is presented in table 3.3. Table 3.3 revealed that at any particular heating temperature for each of the particle size, the observed means of oil yield are significantly different from each other. The implication is that, temperature at 70 °C does not have the same treatment effect on oil yield as temperature at 90 °C and vice versa. The table has also shown that for each particle size, the highest mean value of oil yield occurred at heating temperature of 110 °C, while heating temperature of 70 °C gave the lowest mean value of oil yield. The low oil yield recorded from samples heated at 70 °C could be attributed to insufficient heat treatment of the samples during the heating process. This suggest that heating temperature of 70 °C was inadequate for enough protein coagulation, breakdown of oil cells, reduction of moisture content to an optimum level and reduction of viscosity that allows for free flow of expressed oil. Table 1: Summary of the Result of Effect of Process Conditions on Oil yield Parameter Pressure (MPa) Temperature (°C) Particle size 6 9 12 Fine 7.27 13.25 17.7 70 Coarse 1.41 2.20 7.02 Fine 8.37 9.18 8.6 Oil Yield 90 Coarse 1.78 3.93 2.1 (%) Fine 9.18 8.60 13.3 110 Coarse 3.93 2.10 2.19 Fine 8.60 13.3 17.1 130 Coarse 2.10 2.19 3.72

15 22.90 7.79 13.30 2.19 17.10 3.72 18.70 9.43

Table 2: Summary of Analysis of Variance (ANOVA) of the Effects of Temperature, Pressure and Particle Size on Oil Yield Particle Source of Parameter Size Variation df Ss Ms Fcal Prob. Nigerian Institution of Agricultural Engineers © www.niae.net

42


Oil yield

Fine

Coarse

Temperature(A) Pressure(B) AxB Error Total Temperature(A) Pressure(B) AxB Error Total Temperature(A) Pressure(B) AxB Error

3 3 9 32 47 3 3 9 32 47 3 3 9 32

278.864 2692.530 90.643 25.125 3087.162 227.346 598.956 64.139 14.360 904.801 1522.290 9246.183 625.671 288.891

92.955 897.510 10.071 0.785

118.388 1143 12.827

0.001 0.001 0.001

75.782 199.652 7.127 0.449

168.868 444.892 15.880

0.001 0.001 0.001

507.430 3082.061 69.519 9.028

56.207 341.395 7.701

0.001 0.001 0.001

Table 3: Temperature and Pressure comparison on Oil Yield using New Duncan Multiple Range Test (NDMRT) Particle Size Temperature Oil yield (%) Pressure Oil yield (%) 70 15.26a 6 8.36a b 90 18.20 9 16.63b Fine c 110 21.68 12 21.38c 130 20.23d 15 28.99d a 70 4.61 6 2.30a b 90 5.83 9 5.65b Coarse c 110 10.32 12 9.09c d 130 7.95 15 11.67d Means with the same letters are not significantly different at p≤0.05 using the DNMRT A critical look at table 1 and 3 reveals that for each of the particle size, at higher levels of applied pressure (12.0MPa and 15.0MPA) oil yield increased sharply with an increase in heating temperature from 70 °C to 90 °C, while further increase in the heating temperature beyond 110 °C to 130 °C caused a decrease in oil yield. This trend of relationship between oil yield and heating temperature is in agreement with Akinoso et al (2006), Tunder-Akintunde et al (2001), Tunde – Akintunde (2000), Adeeko and Ajibola (1989), and other researchers that had worked in oil expression too numerous to mention. They all affirmed that increase in heating temperature beyond an optimum value will definitely reduce oil yield in mechanical expression process. The early increase in oil yield from both particle sizes due to increase in heating temperature from 70 °C to 90 °C in this study, was not far from the breaking of oil cells and the coagulation of protein, which resulted to decrease in oil viscosity and moisture content, because of the heat on the ground palm kernel samples. This heating, process had been reported by Olaniyan (2006), Ajibola et al 1990, Ajibola (1989) and Adeeko and Ajibola (1989) among others, to have had substantial increase in oil yield. On the other hand, the decrease in oil yield at heating temperatures beyond 110 °C might be attributed to case hardening resulting from significant moisture loss from the ground palm kernel samples at higher temperatures from 110 °C to 130 °C. 3.2

Effects of Applied Pressure on Oil Yield

In order to determine the differences in the treatment effect of applied pressure that contributed more to the changes in the oil yield, New Duncan Multiple Range Test (NDMRT) was employed and the result of the comparison is as presented in Table 3.


43


From Table 3, the mean value of oil yield was highest at 15.0 MPa applied pressure and the lowest mean value of oil yield was at 6.0 MPa for both the fine and coarse particle sizes. The table also shows that there were significant differences between the values of oil yield at all the levels of applied pressure. The table suggests that oil yield from each of the particle sizes increased rapidly with increasing applied pressure from 6.0 MPa to 12.0 MPa compared with increase in oil yield as a result of increasing applied pressure from 12.0 MPa to 15.0 MPa. This increase in oil yield as the applied pressure increases for both fine and coarse particle sizes may be attributed to the shearing of oil capillaries during the application of pressure in the expression process. On the other hand, the reduction in the increase in oil yield may be as a result of sealing of reasonable number of oil capillaries as the applied pressure was increased from 12.0 MPa to 15.0MPa. This observation is in agreement with the observation of Olaniyan (2006), Tunde Akintunde et al (2001), Ajibola et al (1998) and Faborode and Favier (1997). They ascertained that increasing the applied pressure beyond an optimum value will eventually seal the oil capillaries and consequently affect the oil yield. The velocity of flow of the oil from the oil capillaries to the inter-particle voids and finally through the retaining envelop to the oil collecting pan, can also be used to explain the effect of pressure on oil yield from palm kernel during mechanical oil expression under uniaxial loading. Applied pressure of 6.0 MPa may not have been sufficient to result to free flow of oil through the inter-particle void of the ground palm kernel samples, which implies that the rate of flow of oil will be slow, thereby resulting to low oil yield at this pressure. Meanwhile, higher oil yield was obtained at the applied pressure of 12.0 MPa and 15.0 MPa due to greater forces, which increased the rate of free flow of expelled oil through the oil capillaries and out of the inter-particle voids. 3.3

The Effect of Particle Size on Oil Yield

The independent t-test was used to compare the means of fine and coarse particle sizes. The result is shown in table 4a and 4b. Table 4a: Group Statistics Showing the Effect of Particle Size on Oil Yield Parameter Particle size N Mean Std Deviation Oil Yield

Fine Coarse

48 48

18.8450 7.1779

8.10458 4.38760

Std Error Means 1.16980 0.63330

Table 4b: Independent T-tests showing the Effect of Particle size Oil Yield Size t-test for equality of means Parameter Std Error T Df Sig. (2 tailed) Mean Diff. Difference Fine 8.771 94 0.001 11.66 1.33 Oil Yield Coarse 8.771 72.37 0.001 11.66 1.33 From tables 4a and 4b, the mean of oil yield from fine particle sizes is significantly higher than that of coarse particle sizes at 1% level of significance. The t- statistics confirms that the differences observed in the mean values of oil yield between the fine and coarse particle sizes was not due to random occurrence alone. This significant difference in the oil yield from fine and coarse particle sizes is also evident in table 3.3 as seen earlier. More oils obtained from fine particle sizes as compared to coarse particles sizes can be attributed to the weakening of more oil cell walls during size reduction. The weakened (or broken) oil cell-walls readily Nigerian Institution of Agricultural Engineers © www.niae.net

44


expel oil on the application of little pressure. Also, size reduction exposed a greater area of oil bearing cell walls to heat (during heat treatment) and pressure application. 4.

CONCLUSION

From this study, the effect of heating temperature, applied pressure and particle size on oil yield, during mechanical expression of oil from palm kernel under uniaxial compressive loading was revealed. The following conclusions are drawn: (i) Increase in heating temperature from 70 °C to 110 °C irrespective of the applied pressure, results to a corresponding increase in the oil yield from the fine and coarse particle size, but further increase in heating temperature from 110 °C to 130 °C results in decrease in the oil yield from both particle sizes. (ii) Oil yield increase as the applied pressure increases from 6 MPa to 15 MPa at any heating temperature for the fine and coarse particle size respectively. The highest oil yield was recorded at applied pressure of 15 MPa while the minimum was at 6 MPa. (iii) In order to maximize oil yield, the process conditions (i.e. heating temperature and applied pressure must be properly controlled during the mechanical oil expression process. (iv) After an optimum temperature of 110 °C, there will be a reduction in the oil yield of palm kernel oil mechanically expressed from fine and coarse particle size of palm kernel seed. (v) Oil yield from fine particle size are higher and significantly different from that obtained from the coarse particle size. (vi) Physico-chemical characteristics analysis of the expressed oil showed that the quality of oil obtained reduces to a large extent as the heating temperature increases from 70 °C to 130 °C. REFERENCES Adeeko, K.A., and Ajibola, O.O. 1989. Processing Factors Affecting Yield and Quality of Mechanically Expressed Groundnut Oil. Journal Agricultural Engineering Research. 5:31-43. Ajibola, O.O. 1989. Mechanical Expression of Oil from Melon Seeds. Proceedings of the 11th International Congress on Agricultural Engineering. Dublin, Ireland. 4:2409-2416. Ajibola, O.O., Bakare, F.A., Adeeko, K.A., and Fashina, O.O. 1990. Effects of some Processing Factors on Yield of Oil Expressed from Rubber Seeds.Ife Journal of Technology. 2(2): 17. Akinnoso, R., Igbeka, J.C., Olayanju, T.M.A. and Bankole, L.K. 2006. Modeling of Oil Expression from Palm Kernel. Agricultural Engineering International: the CIGR E-journal Vol. VIII, October. AOAC. 1991. Official Methods of Analysis. Association of Official AnalyticalChemists, Washington DC, USA. Ajibola, O.O., Adeyemo, A.O., and Owolarafe, O.K. 1998. Mechanical Expression of Oil from Castor Seeds. Journal of Agricultural Engineering and Technology. Vol. 6: 1-11. ASAE. 1989. Methods of Determining and Expressing Fineness of Feed Materials by Sieving. ASAE Standards: ASAE S319. Agricultural Engineers’ Handbook. Pp. 325-327. Bamgboye, I.A and Adejumo, O.I. 2011. Effects of Processing Parameters of Roselle Seed on its Oil Yield. Int. J.Agric and Biol. Eng. Vol.4(9): 82 – 86. Breadson, D.K. 1983. Mechanical Extraction. Journal of American Oil Chemists Society. 60(2): 211213. Faborode, M.O. and Favier, J.F. 1997. New Insight into the Mechanics of Seed-Oil Expression. Journal of Agricultural Engineering and Technology (JAET) Vol. 5: 9-24. Hamzat, K.O. and Clerke, A.O. 1993. Prediction of Oil Yield from groundnut using the concept of Quasi –Equilibrium Oil Yield. Journal of Agricultural Engineering Research. Vol. 55(1): 79-87. Hartley, C.W.S. 1988. The Oil Palm. Third edition. John Wiley and Sons Inc. New York. Pp.19 – 21. Mrema, G.C. 1979. Mechanisms of Mechanical Oil Expression from Rapeseed and CashewUnpublished PhD Thesis, National University of Ireland, Dublin, Ireland.


45


Mrema, G.C. and McNutty, P.B. 1980. Mechanisms of Oil Expression from Rapeseed and Cashew. In: Solids Separation Process. Chemical Engineering Symposium, London, United Kingdom. Series 50. Pp. 1-11. Mrema, G.C. and McNulty, P.B. 1984. Microstructure of Rapeseed and Cashew as related to Mechanical Oil Expression. Irish Journal of Food Science and Technology. 8:59-66. Mrema, G.C. and McNulty, P.B. 1985. Mathematical Model of Mechanical Oil Expression from Oil Seeds. Journal of Agricultural Engineering Research. 31:361-370. Olaniyan, A.M. 2006. Development of Dry Extraction Process for Recovering Shea Butter from Shea Kernel. Unpublished Ph.D thesis, Department of Agricultural Engineering, University of Ilorin, Nigeria. Olaniyan, A.M. 2010. Effect of Extraction Conditions on the Yield and Quality of Oil from Castor Bean. Journal of Cereals and Oil Seeds Vol. 1(2):24 – 33. Pantzaris, T.P. and Ahmed M.J. 2001. Properties and Utilization of Palm Kernel Oil. Malaysia Palm Oil Board (Europe) Brickendonbury, Hertford SG138NL, UK. Pantzaris, T.P. 2000. Pocket book of Palm Oil Uses. Fifth Edition, MPOB, Bangi. Pp.102 – 109 Praveen, C.B. 1997. Mechanical Oil Expression from Selected Oil seeds Under Uniaxial Compression. Unpublished Ph.D. thesis submitted to the College of Graduate Studies and Research, Department of Agricultural and Bioresource Engineering, University of Saskatchewan, Canada. Tunde-Akintunde,T.Y., Akintunde, B.O. and Igbeka, J.C. 2001. Effects of Processing Factors in Yield and Quality of Mechanically Expressed Soybean Oil. Journal of Agriculture Engineering and Technology. 9:39-45. Tunde-Akintunde, T.Y. 2000. Predictive Models for Evaluating the Effect of some Processing Parameters on Yield and Quality of some Soybean (Galycine max (1) merri) Products. PhD Thesis, Department of Agricultural Engineering, University of Ibadan, Ibadan, Nigeria. Young, J.H.; Whitaker, T.B.; Blankenshy, P.D., Brusewitz, G.H. and Steele, J.C. 1982. Effect of OvenDrying Time on Peanut Moisture Content Determination. Transactions of the American Society of Agricultural Engineers. 25(2):4-9.


46


DEVELOPMENT OF A SOLAR CABINET DRYER FOR ROOT CROPS CHIPS IN NIGERIA C. O. Nwajinka1 and C. U. Onuegbu2 Department of Agricultural and Bioresource Engineering, Nnamdi Azikiwe University, Awka, Nigeria 2 Department of Agricultural Engineering, Federal Polytechnic, Oko, Nigeria. Email: [email protected] 1

ABSTRACT A solar cabinet dryer was developed and tested using cassava chips. The test involved drying experiments with fresh cassava chips of average size of 3.5 cm. In the trial experiment, the dryer was not loaded with the product rather, temperature of the collector, the dying chamber/plenum and chimney were taken at intervals of 30 minutes. The highest temperature attained by the solar dryer was 45.5°C. A sample of 200g of cassava chips was thereafter loaded in each of the three trays and dried from moisture content of 80% (wb) to 14.3 % (wb) in tray 1, to 15.3% (wb) in tray 2 and to 14.3% (wb) in tray 3 respectively in four days. The quality deterioration of the dried cassava chips based on colour change was not significant by visual observation. This indicates that there was little or no fungal (mould) infestation and is therefore recommended for use for cassava chips drying in Nigeria if similar tests are carried out in other agroclimate zones of the country. A performance test for other sizes of the chips is ongoing. KEYWORDS: Solar cabinet dryer, cassava chips, root crops. 1.

INTRODUCTION

Sun-drying method may be efficient and cheap process but has disadvantages such as contamination, insects and bacteria infestation and loss due to wetting by rain squalls. In order to protect the products from above mentioned disadvantages and also to accelerate the time for drying the products, reduce the moisture and hence wastage through bacterial action, different types of solar dryer has been developed (Excel 1980, Janjaia et al 2008, Yaldyz and Ertekyns 2001). During the mid 1970’s shortages of oil and natural gas, increase in the cost of fossil fuels and the depletion of other fossil fuel resources stimulated efforts in the development of solar energy as a practical power source. Thus, interest was enkindled in the harnessing of solar energy for heating, cooling, generation of electricity and other purposes. It also found applications in the area of agriculture for crop drying, irrigation pumping, and numerous other thermal processes in food industries. Cassava is one of the most widely used foods in Nigeria. Its storage poses big problem to farmers and is traditionally left in the soil until it is set to be used. It has high moisture content and high respiration rate thereby subjecting it to microbial, mechanical and physiological damages after harvest. All these factors lead to high post-harvest losses, thereby necessitating drying of the portion of the production that will not be readily consumed. However, drying faces a big challenge because of the rising energy cost and the global disapproval of the conventional fossil fuels as sources of energy. The basic essence of drying is to reduce the moisture content of the product to a level that prevents deterioration after harvest (Rajkumar, Kulanthaisami et al. 2006). In many parts of the world there is a growing awareness that renewable energy sources have an important role to play in extending technology to the farmer in developing countries to increase their productivity (Waewsak et al., 2006). Solar thermal technology is rapidly gaining acceptance as an energy saving measure in agricultural application. It is preferred to other sources of energy because it is abundant, inexhaustible and non-polluting (Akinola, 1999; Akinola and Fapetu, 2006; Akinola et al., 2006). Solar air heaters are simple devices that raise the temperature of air by solar energy and are employed in many applications requiring low to moderate temperature below 800C, such as crop drying and space heating (Kurtbas and Turgut, 2006).


47


The post-harvest loss of agricultural products in most developing countries is enormous. This is a serious concern for an agricultural country like Nigeria where approximately 91% are dependent upon subsistence farming for their survival. As a consequence, farmers income remain low due to low farm gate prices and retail prices remain high as the losses are passed on to consumers. However, several factors contribute to post-harvest losses and some of the technological factors include faulty harvesting and handling practices, poor packaging and transport systems, lack of storage facilities and poor processing techniques. Therefore, the food processing sector can play a vital role in reducing the post-harvest losses through processing and value addition. The objective of this work, therefore, is to develop a solar dryer that will serve the purpose of the rural farmer to address the problems of preservation of such high moisture content crops like fruits, vegetables and tubers. 2.

MATERIALS AND METHOD

2.1

Description of the Solar Food Dryer

The solar food dryer consists of two major integral components (Figs 1 & 2) namely: (i.) The solar collector compartment, which can also be referred to as the air heater. (ii) The drying chamber, designed to accommodate layers of drying trays made of chicken mesh on which the produces or food are placed for drying The fan was powered by a solar panel connected to a controller unit.

Fig. 1: The Solar Cabinet Dryer


48


Fig 2: The pictorial view of the Solar Cabinet Dryer 2.2

Basic Design Calculations

The size of the dryer was determined as a function of the drying area needed per kilogram of cassava chips. Calculation of the Amount of Moisture to be removed: The amount of moisture to be removed from the product, Mw in Kg is estimated from the following expression. 𝑀𝑤 =

𝑀𝑔 (𝑀𝑖 −𝑀𝑓 ) 100−𝑀𝑓

(1)

Where: Mg = initial mass of wet crop to be dried (kg); moisture content (%).

Mi = initial moisture content (%); Mf = final

The Heat Gained by the Collector per Unit Time Qc: This is given by; Qc = A [I - UL (Ti – Ta)] …..

(2)

Where: A = area of the transparent cover (M2) I = incident insolation (w/m2) UL = overall heat loss for the collector (w/c)  = solar absorbance = transmittance  Ti = temperature of incoming air Ta = temperature of ambient air


49


Since the collector draws the ambient air directly and the length is not too long, the last term on the righthand side vanishes and the rate of energy collection is simply estimated using; Qc

=

AI ….. (3)

(3) If the mass of air leaving the collector per unit time is ṁa, the heat gained by the air Qa is;

Where:

Qa C To

= = =

ṁaC (To – Ti) …. (4) specific heat capacity of air (KJ/kg‾1˚C) temperature of out-going air

(4) A simplified energy equation for the dryer is Qc = Qa i.e AI

=

ṁaC (To – Ti) …. (5)

Therefore, the required surface area of the transparent cover, which determines the size and dimensions of the dryer, is obtained from; 𝐴𝐶 =

𝑚̇𝑎 𝐶(𝑇𝑜 − 𝑇𝑖 ) 𝐼𝜏𝛼

(5) The total energy required for drying a given quantity of food items can be estimated using the basic energy balance equation for the evaporation of water; ṁwLv =

ṁaC (To – Ti) …. (7)

Where LV ṁw

= specific latent heat of vaporization of water from the food = mass of water evaporation from the food item (kg/s)

surface (kJ/kg)

The mass of water mw is estimated from the initial content Mi and the final desired moisture content Mf as stated in equation (1) as 𝑚̇𝑤 =

𝑀𝑔 (𝑀𝑖 − 𝑀𝑓 ) 100 − 𝑀𝑓

(6) The Average drying rate is given by 𝑀𝑑𝑟 =

𝑀𝑤 𝑡𝑑

Where: td = total drying time Mdr = Average drying rate (kg/hr) Mw = Amount of water/moisture to be removed (7) The Quantity of Air Needed for Drying The quantity of air needed for drying may be estimated from the energy balance equation. The basic energy balance equation for drying process is MwL = Ma La a (Ti – Tf) …… (9)


50


𝑀𝑎 =

𝑀𝑤 𝐿 𝐶𝑎 𝜌𝑎 (𝑇𝑖 − 𝑇𝑓 )

Where: Ma = Quantity of air needed to absorb Mw kg of water (m3) L = specific latent heat of vaporization of water from the crop to be dried. Ca = specific heat capacity of the air at constant pressure (kJ/kg0C). a = Density of drying air (kg/m3) Ti and Tf = The initial and final temperature of the drying air before passing through the drying bed (0C). (8) Quantity of Heat needed to evaporate the water. The quantity of heat required to evaporate the water would be …. (11)

Q = Mw .hfg

Where: Q = The amount of energy required for the drying Process KJ Mw = mass of water (kg) hfg = latent heat of evaporation of water (KJ/kg) 2.3

Design Consideration and Parameters 1. Angle of Tilt (β) of solar collector (Air heater). Duffie and Beckman, (1974) suggested that the angle of tilt (β) of the solar collector can be estimated by adding 10o to the Latitude of the location. Therefore, β

=

100 + Lat Ø

Were latitude Ø is the latitude of the collector location, the latitude of Awka where the drying was designed is latitude 6.200. Hence, the suitable value of β use for the collector. β

=

100 + 6.200 = 16. 200

2. Isolation on the collector surface Area. The average daily solar radiation for Awka is 5157.22kJ/m2/day to 53002.94kJ/m2/day (Okonkwo and Nwokoye, 2011). 3. Air gap – A gap of 5cm is created as air vent (inlet) and air passage. 4. Glass and flat plate collector – transparent glass covering of 60 x 60cm2 and 4mm thickness glass was used. 5. Drying chamber was of average dimension of 60 x 57 x 55cm with chimney 6. Drying Trays: Chicken mesh was selected as the dryer screen or trays to aid air circulation within the drying chamber. Three trays each having dimensions of 50 x 50cm. 2.4

Experimental Procedure

The experiment was conducted in Nnamdi Azikiwe University Awka located at longitude 7.12˚E and latitude 6.20˚N from 8th April to 12th May 2011. Experiments were carried out using cassava chips. The cassava chips were prepared, washed, weighed before drying. The chips were left inside the drying chamber and monitored for weight loss at regular intervals. Temperatures of the drying medium and ambient were measured at one hour intervals. 100g of the cassava chips cut in circular shape was placed on each of the three trays in the drying chamber.


51


The air temperature at collector inlet, collector temperature, and surface temperature of the drying trays were measured using laboratory type digital thermometer (accuracy ± 0.5˚C ) between the hours of 9.00am and 4.00pm each day. The weight of the cassava chips were taken periodically every thirty minute until no further weight loss was noticed. Initial weight as well as the final weight was noted. The weight loss was used to calculate the moisture ratio and drying rate at 30 minute intervals as the chips dried. 2.5

Determination of Moisture Content

The moisture content of the cassava chips was determined using oven method. Chips from the same samples to be dried were weighed in moisture cans with known weights. The sample and tray were put in an oven at temperature of 103˚C. After intervals of 12 hours, the tray plus sample was weighed to note the change in weight, and was put back to the oven. The process of weighing and returning to oven continued until there was a constant weight. Thus: (Wi − Wf )100 % Moisture Content (wet basis) = (Wi ) Where, Wi=initial weight of sample, Wf=final weight of sample The moisture content was determined before and after each experiment to represent the initial and final moisture contents respectively. 3.


The temperature profile of the dryer was determined by measuring the temperature inside the drying chamber and the ambient air hourly, between 8.am to 4.pm (local time). Moisture removed from cassava chips was monitored hourly as the weight during the day. The results are presented in Tables 1 – 3 and Figures 1 – 5.

Dryer Wet bulb tempt T (0C)

Ambient Wet bulb tempt T (0C)

Ambient Relative humidity (%)

Dryer Relative humidity (%)

Chimney Tempt T(0C)

27 28 29 30 34.7 35.7 29 31

Solar collector Tempt T (0C)

Tray2 surface tempt T (˚C)

29 29 29.8 30 34 35.2 30 32

Ambient Tempt T0 (0C)


9.am 10.am 11.am 12.pm 1.pm 2.pm 3.pm 4.pm


Local Time (hr)

Table 1: Result of no-load test run of the dryer (without cassava chips) in August 2012 for (Day 1)

28 29 30 31 32 34 31 30

28 27 28 28 33.6 32.2 27 28

34 31 32 33 41 45 31 32

26 25 26 25 28 30 29 27

25 24 25 24 26 27 26 24

80.3 79.5 80.3 73.7 58.4 70.3 93.2 73.7

56.7 64.4 65.5 55.4 41.6 38.6 88.1 71.2

30 31 29 28.5 27 30,5 32 31


52







Ambient wet bulb tempt T (0C)


Relative humidity of drying chamber (%)

Chimney Tempt T(0C)

10.am 11.am 12.pm 1.pm 2.pm 3.pm 4.pm


Local Time (hr)

Table 2: Results of experimental drying of cassava chips and dryer parameters in August 2012 (Day 2)

32.0 31.5 34.0 38.0 34.0 34.5 32.0

30 30.5 32.0 32.0 33.0 30.5 30.0

36.0 32.0 36.0 40.0 39.0 38.0 34.0

30.0 28.0 32.0 33.0 32.0 30.5 3O.0

40.0 32.0 40.0 38.0 34.0 34.0 32.0

24.0 26.0 34.0 28.0 28,5 27.5 27.0

28.0 24.0 30.0 30.0 30.0 28.0 26.0

87.7 73.7 88.4 83.3 88.5 84.9 75.5

26.4 65.5 72.4 51.6 70,2 64.8 71.2

31.0 31.5 34.0 32.0 35.0 34.0 30.0

Local Time (hr)

Tray 1 surface tempt T (˚C)






Ambient Wet bulb tempt T (0C)


Dryer Relative humidity (%)

Chimney Tempt T(0C)

Table 3: Results of experimental drying of cassava chips and dryer parameters in August 2012 (Day 3)

10.am 11.am 12.pm 1.pm 2.pm 3.pm 4.pm

32.0 34.0 34.0 34.0 36.0 34.0 34.0

30.0 30.0 30.1 32.0 32.5 32.0 32.0

32.0 38.0 36.0 38.0 40.0 40.0 38.0

31.0 32.0 30.2 32.0 32.0 32.0 30.0

34.0 40.0 32.0 34.0 36.0 36.0 34.0

31.0 30.0 26.0 28.0 29.0 30.0 28.0

30.0 28.0 28.0 27.0 28.0 29.0 27.0

94.1 77.0 92.7 77.0 77.0 88.5 87.7

83.8 54.0 65.5 62.1 64.2 64.2 67.5

28.0 35.0 34.5 35.0 36.0 38.0 34.0

The solar radiation was discovered to fluctuate with time as shown in the plot of average daily solar radiation against the day of the month (figure 1).


53


Daily Solar radiation (kJ/m2/day)

Solar radiation, (kJ/m2

60000 50000 40000 30000 20000 10000 0 0

10

20

30

40

Day

Solar radiation, (kJ/m2/day-1)

Figure 1: Daily hourly radiation at Awka for the month of April, 2011

Daily Solar radiation, (kJ/m2day-1) 50000 40000 30000 20000 10000 0 0

10

20

30

40

Day of the Month Figure 2: Daily hourly radiation at Awka for the month of May, 2011 As the time progresses, the dryer temperature increases until around 12 noon and 1.00pm local time for the ambient and dryer respectively. The ambient peaked first before the drying chamber, by a time lag of one hour. The temperature inside the dryer was higher than the ambient air temperature throughout the greater part of the day for the period tested. This indicates prospect for higher rate of moisture removal in solar dryer than open–air sun drying.


54


Fig.3: Average ambient and collector temperatures against local time for the month of April 2011

50 Tray surface…

Temperature, oC

45 40 35 30 25 20 0

5

10 (hrs) 15 Local time,

20

Fig.4: Average Tray temperatures against local time for the month of April 2011 The drying was quite slow in the morning as the drying chamber warmed up and the solar intensity low. The maximum drying occurred between 12.pm - 1.pm local time which corresponds with the period of highest temperature inside the dryer and high solar radiation intensity. As the solar radiation increases, the temperature on the collector and inside the dryer increases. This logically indicates higher rate of moisture transport in solar dryer during the mid-day


55


Fig 5: Average temperatures of the three drying trays against local time Relative humidity was calculated with the values of wet bulb and dry bulb temperatures using the Vaisala sychrometric calculator. The results are presented in Tables1 to 3. 4.

CONCLUSIONS

The rate of moisture removal from food items is a function of the temperature inside the dryer. Solar radiation can be effectively utilized for drying of agricultural product in Awka environment. Locally available materials were used in construction making it available and easy to maintain especially by peasant farmers. This will go a long way in reducing food wastage and consequently food hortages. This dryer can be used extensively for majority of the agricultural food crops. It protects the environment and saves cost and time spent on open sun drying of agricultural produce. The food items are also well protected in the solar dryer than in the open sun, thus minimizing the cases of pest, insect attack and contamination. However, the performance of the solar food dryer can still be improved upon especially in aspect of reducing the drying time and probably storage of heat energy within the system. REFERENCES Akinola, A.O. 1999. Development and Performance Evaluation of a Mixed-Mode Solar Food Dryer. M. Eng. Thesis, Federal University of Technology, Akure, Nigeria. Akinola, A.O.; Akinyemi, A.A; and Bolaji, B.O. 2006. Evaluation of traditional and solar fish drying systems towards enhancing fish storage and preservation in Nigeria. J. Fish. Int., Pakistan 1 (3-4): 449. Akinola, A.O; and Fapetu, O.P. 2006. Exergetic Analysis of a Mixed-Mode Solar Dryer. J. Engin. Appl. Sci. I: 205-10. Ayensu, A., Optimizing the performance of solar dryers, Ghana Journal of Energy Research &Technology (ERG Bulletin), 1994, Vol.6, pp. 119 - 130. Bassey, M.W. 1989, Development and use of solar drying technologies Nigerian Journal of Solar Energy 89: 133-64. Duffie, J. A. and Beckman, W. A., 1991, Solar Engineering of Thermal Processes, John Wiley & Sons Inc., New York. Ertekin, C.; and Yaldiz, O. 2004. Drying of eggplant and selection of a suitable thin layer drying model, J. Food Engin. 63: 349-59. Exell, R.H.B., 1980. Basic design theory for a simple solar rice dryer. Renewable Energy Review Journal 1 (2), 1–14. Farkas.; Scalin D., 2004 The Design, Construction and Use of an Indirect, Through-pass, Solar Food Dryer, Home Power Magazine, 1997, 57, p. 62-72. Nigerian Institution of Agricultural Engineers © www.niae.net

56


Ikejiofor, I.D. 1985. Passive solar cabinet dryer for drying agricultural products. In: O. Awe (Editor), African Union of Physics. Proc. Workshop Phys. Tech. Solar Energy Convers, University of Ibadan, Nigeria, pp. 157-65. Janjai .S , T. Keawprasert, C. Chaichoet, P. Intawee, B.K. Bala, and W. Muhlbauer 2004. Simulation Model of a PV-Ventilated System for a Solar Dryer. Technical Digest of the International PVSEC-14, Bangkok, Thailand, (2004) P-175. Kurtbas, I.; and Turgut, E. 2006 Experimental investigation of solar air heater with free and fixed fins: efficiency and energy loss. Int. J. Science. Technology. 1(1): 75-82. Kingsly, R. P., Goyal, R. K., Manikantan, M.R., and Ilyas, S.M. 2007. Effects of pretreatments and drying air temperature on drying behaviour of peach slice. International Journal of Food Science and Technology, 4: 65–69. Madhlopa, A.; Jones, S.A.; and Kalenga-Saka, J.D. 2002 A solar air heater with composite absorber systems for food dehydration. Renewable Energy 27: 27-37. Okonkwo and Nwokoye 2011. Measurement and Performance Analysis of Daily Average Solar Radiation at Awka, Nigeria. Journal of Basic Physical Research Vol. 2, No.2, pp 7 - 13, Available online at www.jbasicphyres-unizik.org Saeed. I.E, K. Sopian and Z. Zainol Abidin. Thin-Layer Drying of Roselle (I): Mathematical Modeling and Drying Experiments. Agricultural Engineering International: the CIGR Ejournal. Manuscript FP 08 015. Vol. X. September, 2008. Togrul, I.T.; and Pehlivan, D. 2004. Modeling of thin layer drying kinetics of some fruits under open-air sun drying process. J. Food Engineering 65: 413-25. Tunde-Akintunde T.Y. and Afon A.A. 2009. Modelling of Hot-Air Drying of Pretreated Cassava Chips Agricultural Engineering International: the CIGR Ejournal. Manuscript 1493 Vol. August Waewsak, J.; Chindaruksa, S.; and Punlek, C. 2006. A Mathematics modeling study of hot air drying for some agricultural products. Thammasat Int. J. Science Technology 11(1): 14-20. Yaldiz, O. and C. Ertekin, 2004. Drying of egg plant and selection of a suitable thin layer drying model. J. Food Eng., 63: 349-359. Appendix Table 1: Summary of Design materials and parameters. S/N Parameters 1 Location 2 Drying period 3 Solar irradiation, Awka (October/ November) 4 Average relative humidity 5 Food products 6 Initial moisture content in yam 7 Mode of heating 8 Number of glazing 9 Glazing slope 10 Glazing Materials 11 Loading provision 12 Number of trays 13 Air outlet provision 14 Air circulation mode Forced convection 15 Drying capacity 16 Thickness of plantain chips 17 Construction materials 18 Insulation used 19 Thickness of glass 20 Transmittance of glass 21 Emissivity of cover Nigerian Institution of Agricultural Engineers © www.niae.net

Description Awkas, ( lat 6.20 o’N) April- May, 2011 15.4 – 16.7 MJ/m2/day 89% Plantain and Cassava 65-85% Indirect 1 17o Glass Door at back side of chamber 3 Vent at top of the chimney 10kg 3 mm Wood, glass, mild steel sheet Glass fibre 4.5 mm .81 0.88 57


22 23 24 25 26 27 28 29 30

Emissivity of plate Sky temperature Air velocity at fan speed of 750 rpm Dimension of collector Insulator thickness Pv module type and rating Solar charge controller Fan rating Fan speed


0.98 32 oC 5 m/s 1.08 x 0.76 m2 0.0125 m 55Wp (RSM 50, Solar smart switch) SDRC- 10IP (12/24 V auto) 12 Volts 702 - 1012 rpm

58


EFFECT OF DRYING METHODS AND PRODUCTION PROCESS ON THE QUALITY PARAMETERS OF UNFERMENTED CASSAVA FLOUR C. A. Adamade, B. A Jackson and F. Agaja Department of Agro-Industrial Development and Extension Department National Centre for Agricultural Mechanization, (NCAM), Ilorin, Nigeria. Email: [email protected] ABSTRACT High Quality Unfermented Cassava Flour is a product of cassava root that is acceptable with high market and nutritional value. Two different drying methods were tested to determine its quality requirements and the effect of the drying methods used. Investigations into the effects of drying methods and production process on the quality parameters of Unfermented Cassava Flour revealed that there is a significant difference in the flour composition by percent dry weight. The quality parameters investigated were Moisture Content, Protein Content, Ash Content, Fat Content, Carbohydrate Content and Fibre Content. Results obtained showed that flour obtained using NCAM kerosene fired batch dryer (Sample A) has a mean Moisture Content 10.03%, Protein Content 0.14%, Ash Content 1.08%, Fat Content 0.6%, Fibre Content 1.69%, Carbohydrate Content 79.85% and HCN 14.03 mg/kg while cassava flour obtained using sun drying (Sample B) has a mean moisture Content 11.64%, Protein Content 0.13%, Ash Content 4.47%, Fat Content 2.06%, Fibre Content 3.85%, Carbohydrate Content 80.33% and HCN 14.62 mg/kg. The result of product from NCAM kerosene fired batch dryer conforms to the Nigerian Industrial Standard. This indicates that NCAM kerosene fired dryer dries crop faster and performs better when compared to the sun drying method. KEYWORDS: Cassava flour, drying, unfermented, ash content, fat content, protein content. 1.

INTRODUCTION

Cassava (Manhot species) is a crop which tolerates drought and low soil fertility and is primarily grown by small scale farmers in area with marginal soils or unfavourable climate. It has the ability to withstand poor environmental condition (Nweke et al 2000). It has now become a major economic activity in Nigeria due to the encouragement given by the government through Root and Tuber expansion programme. According to Ihekoronye and Ngoody, (1983), Cassava root constitute the following: Starch 20 – 30 %; Protein 2 – 3 %; Water 75 – 80 %; Fat content 0.1 % and Ash content 1 – 1.5%. The processing of cassava roots into cassava flour involves the following steps: washing, peeling, washing, grating, pressing, drying, milling and sieving for unfermented cassava flour. This process does not act on the cyanogenic potential (Hahn, 1989) and is suitable for the sweet cassava varieties. On the average cassava flour contains as stated by Bamiyo (2003): Moisture Content Protein Carbohydrate Ash Fibre Particle size

10% 1.5% 86% 0.9% 2.0% 350 – 350 micro

Traditionally, cassava flour known as lafun is a product of dried fermented or unfermented cassava chips or mash which is then milled to desired particle sizes. There are two basic process lines required for the production of unfermented cassava flour: chips and grated mash. The resultant flour in this process is known to be white in appearance (Bamiyo, 2003). Cassava flour is often made at home from unfermented cassava chips, so chips are an important intermediate product making up a major proportion of marketed Nigerian Institution of Agricultural Engineers © www.niae.net

59


cassava in Africa (Hahn 1989). Drying is necessary to improve the quality, market value and to preserve agricultural products from spoilage during storage. The elimination of total cyanide from cassava chips are done by drying with heated air. Cassava roots are dried by sun drying for 2 to 5 days depending on the weather and by using artificial drying method (Batch dryer) for some hours. However, cassava has its problem. It is full of carbohydrate and very little protein. On digestion, carcinogenic glycosides present in the root are broken down and cyanide can be released into the body system. These can be removed by washing the cassava in clean water and dry. It is estimated that 40% of the cassava production in Nigeria is lost due to lack of processing capacity and market outlet (Nweke et al 1998). Therefore, commercial production of unfermented cassava flour will bring about rapid socioeconomic transformation, which will raise the standard of living of many rural dwellers, create job opportunities enhances its marketability and also assist in the production of animal feeds for livestock (through pellets). This paper therefore examines the effect of two drying methods on the quality of cassava flour obtained through chips. The objectives of this study are to: determine the effect of each of the drying methods on the quality of unfermented cassava flour; compare the quality parameters of flour obtained from the two drying methods and determine the most appropriate drying method that will produce high quality unfermented cassava flour. 2.


2.1

Description of the Dryer

The batch dryer of dimension 2501 mm by 1165 mm by 810 mm was designed and fabricated at NCAM. It consists of a blower unit, belt pulley transmission system, burner unit, 4.5 KW diesel engine, the plenum and drying tray of dimension 2 mm diameter. The dryer was constructed with 1.5 mm galvanized sheet and 50 mm by 50 mm angle iron. 2.2


The research was conducted at the National Centre for Agricultural Mechanization (NCAM) Ilorin. Fresh Sweet Cassava roots (Manihot Palmata) were purchased from the market, sorted, peeled and washed with clean water and allowed to drain. The roots were then converted into chips using NCAM Chipping machine powered by a 5Hp prime mover. The mash was dewatered with a screw press designed and fabricated in NCAM. Drying operations were done using NCAM designed kerosene fired batch dryer and sun drying respectively. Samples were milled using NCAM designed milling machine operated by a 7hp prime mover. The flowchart for the process is shown in Fig 1. The analysis was carried out using AOAC (2000) method at the University of Ilorin Chemistry Laboratory. The factors investigated include: moisture content, ash content, protein content, fibre content, fat content and carbohydrate content. i. Moisture content: This was determined using oven drying method ii. Protein content: Determined using the Micro Kjeldah method iii. Ash content: Calculated using 100 × xy W Where x = weight of crucible + ash; y = weight of crucible; w = weight of sample (g) before ashing iv. Fat content: Calculated using A – B x 100 C Where A = Initial weight of sample + thimble; B = Final weight of sample + thimble; C = Initial weight of sample


60


v. Fibre Content: Calculated using

𝐶1 −𝐶2 𝑊

𝑋 100

Where C1 – C2 is weight of fibre; W is weight of sample vi. Carbohydrate Content: The total percentage carbohydrate content was determined by difference method. This involves adding the total values of protein, fat, ash, fibre and moisture content and subtracting it from 100. vii. Cyanide Content: The Cyanogenic glycosides (HCN) content was determined by the Argenimetric titration method.

Harvest/Sorting of Cassava root

Peel and wash

Chipping

NCAM Kerosene Fired Batch Dryer

Sun Drying

Milling Milling

Cassava flour Cassava flour Fig. 1. Process flowchart for the production of the cassava flour 3.


3.1

Proximate Composition/Cyanide Content of the Flour

Table 1 shows the experimental values for cassava flour processed by using NCAM Kerosene Fired Batch Dryer, Table 2 shows the experimental values for cassava flour processed by using Sun-drying method and Table 3 indicates the mean values for both drying methods. Analysis with student T test revealed that there is significant difference at 5% confidence level of the products from each of the drying method. 3.2

Effect of Moisture Content on the Drying Methods

Results obtained from using NCAM Kerosene fired dryer (Sample A) showed a Moisture Content of 10.03% which compares favourably with the specification by the Nigerian industrial standard (NIS, 2004). Ogazi and Jones (1990) also reported that 10% moisture content in flour is ideal for good keeping quality while Moisture Content for Sample B showed 11.64% which is above recommended level. The Nigerian Institution of Agricultural Engineers © www.niae.net

61


independent T-test group Statistics which was carried out to compare the means of the two different variables of interest; NCAM batch drying and Sun drying method, with assumption of equal variance showed that the mean values of moisture content from the two drying methods are significantly different from each other at 5% level of significant. Furthermore, the descriptive statistics of the two methods and standard deviation of the moisture content from the two factors showed that the mean value of 11.6360 and the standard deviation of 0.01517 for sun drying tend to behave differently to mean value (10.0140) and standard deviation (0.02702) of NCAM batch drying method. 3.3

Effect of Fibre Content on the Drying Methods

The crude fibre for Sample A was 1.69% which is in line with Sanni et al, (2003). The maximum value stated by NIS (2004) prescribed 2.0% as the maximum. For Sample B, the crude fibre recorded 1.86% which is also within the range. The independent T-test group Statistics was carried out to compare the means of the two different variables of interest, NCAM batch drying and Sun drying method, with assumption of equal variance. From the result, mean value of Fibre content from the two drying methods were significantly different from each other at 5% level of significant. Also the mean value (1.8580) and the standard deviation (0.00837) of Fibre content from sun drying tend to behave differently to mean value (1.6860) and standard deviation (0.00548) of NCAM batch drying method. Table 1: Experimental Value of Analysis on Product from Batch Drying Rarameters 1 2 3 Ash Content (%) 0.89 0.90 0.89 Moisture Content (%) 10.03 10.02 10.04 Protein Content (%) 0.14 0.14 0.14 Fat Content (%) 1.08 1.10 1.09 Carbohydrate Content (%) 79.85 80.00 79.90 Fibre Content (%) 1.69 1.68 1.69 HCN mg/kg 14.02 14.01 14.01

4 0.75 9.97 0.13 1.10 79.99 1.69 14.00

5 0.80 10.01 0.14 1.08 80.01 1.68 14.02

Table 2: Experimental Value of Analysis on Product from Sun Drying Rarameters 1 2 3 Ash Content (%) 1.85 1.87 1.86 Moisture Content (%) 11.62 11.65 11.64 Protein Content (%) 0.128 0.25 0.13 Fat Content (%) 1.05 1.06 1.05 Carbohydrate Content (%) 80.33 80.36 80.37 Fibre Content (%) 1.87 1.85 1.86 HCN mg/kg 14.60 14.62 14.63

4 1.87 11.62 0.121 1.08 79.9 1.86 14.60

5 1.90 11.65 0.127 1.06 80.20 1.85 14.59

Table 3 Rarameters Ash Content (%) Moisture Content (%) Protein Content (%) Fat Content (%) Carbohydrate Content (%) Fibre Content (%) HCN mg/kg 3.4

NCAM – Fried Dryer 1.08 10.03 0.14 1.6 79.85 1.69 14.03

Sun-drying 1.05 11.64 0.13 1.05 80.33 1.86 14.61

Effect of Fat Content on the Drying Methods

The Fat content for sample A was 0.6% which if compared with Sanni et al (2003) had a maximum value of 0.6% while for Sample B was higher than the maximum value stated. The Ash Content obtained for Nigerian Institution of Agricultural Engineers © www.niae.net

62


Sample A was 1.08% and Sample B was 1.05 %. Apea et al, (2007) gave an ash content of 0.70% 2.21%. Despite the differences the products contain some mineral content and thus can be used for animal feed production. The independent T-test group Statistics was carried out to compare the means of the two different variables of interest, NCAM batch drying and Sun drying method, with assumption of equal variance assumed. The mean values of ash content from the two drying methods are significantly different from each other at 5% level of significant. The statistics of the two methods shows mean and standard deviation of the ash content from the two factors. At the end, the mean value 1.8700 and the standard deviation 0 .01871 of ash content from sun drying tend to behave differently to mean value 0.8460 and standard deviation 0.06731 of NCAM batch drying method. 3.5

Effect of Protein Content on the Drying Methods

The protein content showed 0.14% for Sample A while Sample B was 0.13% which can be as a result of the variety of cassava used. The independent T-test group statistics was carried out to compare the means of the two different variables of interest (NCAM batch drying and Sun drying method). From the test, the mean value of protein content from the two drying methods was significantly different from each other at 5% level of significant. Furthermore, the display shows mean and standard deviation of the protein content from the two factors. The implication of this is that the mean value (0.1520) and the standard deviation (0.05495) of protein content from sun drying tend to behave differently to mean value (0.1380) and standard deviation (0.00447) of NCAM batch drying method. 3.6

Effect of Carbohydrate Content on the Drying Methods

Carbohydrate Content of approximately 80% was obtained for Sample A. This is in line with NIS 2004, which shows that it is full of carbohydrate as a source of energy. The HCN content of 14.03 mg/kg was obtained for sample A. This is within the range specified by Standard Organization of Nigeria for Edible Cassava Flour with maximum of 20 mg/kg. The independent T-test group statistics was carried out to compare the means of the two different variables of interest, NCAM batch drying and Sun drying method, with assumption of equal variance, the analysis carried out showed that the mean value of carbohydrate content from the two drying methods are significantly different from each other at 5% level of significant. The result shows mean and standard deviation of the carbohydrate content from the two factors. This means that the mean value (80.2320) and the standard deviation (0.19766) of carbohydrate content from sun drying tend to behave differently to mean value (79.9500) and standard deviation (0.07106) of NCAM batch drying method. 3.7

Production Process

Drying process which is a conventional production process in chip production was reduced as related to the duration of processing using NCAM kerosene fired batch dryer. Using NCAM batch dryer, chips were dried for approximately 8 hours of processing time while in the case of sun-drying a maximum of 3 days was used. The flour processed using the NCAM dryer has composition that conform with Nigerian Industrial Standard. 4.

CONCLUSIONS

Results from the composition of the cassava flour produced using the two drying methods indicated that: i) The drying methods had effect on the proximate composition. ii) Due to the reduction on drying time and energy for the process, drying of cassava chips with NCAM kerosene fire batch dryer is faster and most satisfactory, than sun drying. iii) Cassava flour produced from both drying methods is recommended for animal feeds because of the high fibre and carbohydrate level.


63


REFERENCES AOAC 2000. Association of Official Analytical Chemists Official Methods of Analysis; D.C. Washington USA. Bamiyo, Festus 2003. Cassava Flour: Profitable Option in Cassava Processing. The Guardian, Saturday, August 23, 2003 Pg B11 Hahn, S. K. 1989. An Overview of African Traditional Cassava Processing and Utilization. Outlook on Agriculture 18:110–118 Ihekeronye, A. I. and Ngoddy, P. O. 1985. Integrated Food Science and Technology for the Tropic. Macmillan Pub. Ltd. London. Pages 10 – 26. IITA 2008. Research to Nourish Africa; Research for Development of Root and Tuber System. Acessed on the 20th June 2008 @ www.iita.org. NIS 2004. Nigerian Industrial Standard NIS 344:2004 Standard for Edible Cassava Flour Nweke F. I., Kapinga R. E., Dixon A. G. O., Ugwu B. O., Ajobo and Asadu C. L. 1988. Production Prospects for Cassava in Tanzania. Collaborative Study of Cassava in Africa (COSCA) Working Paper No. 16, COSCA, IITA, Ibadan, Nigeria. 175 pp. Nweke F. I., Lutete D., Dixon A. G. O., Ugwu B. O., Ajobo, Kalombo N., and Bukaka, B. 2000. Cassava Production and Processing in the Democratic Republic of Congo, Collaborative Study of Cassava in Africa (COSCA) Working Paper No. 22 IITA, Ibadan, Nigeria


64


DEVELOPMENT OF A PORTABLE AIR FLOW DIGITAL METER FOR GRAIN DRYING A. B. Istifanus and C. C. Mbajiorgu Department of Agricultural and Bioresources Engineering, University of Nigeria, Nsukka, Nigeria. E-mail: [email protected] ABSTRACT Technological progress allows more and more instruments to be developed for different purposes based on the prevailing need. The development of a portable air flow digital meter relied on a computer based design. Assembly language was used in writing a set of instructions that were programmed into the micro controller component of the system. This produced an interface which enabled interaction using a monitor. The codes were translated from analogue to digital using a Digital Converter (ADC) and then interpreted in a Liquid Crystal Display (LCD) in m/s. The speed of the fan or the position of the meshlike tray regulated how moist the grain is or how fast the drying exercise is to take place. When the speed of the fan was increased, more air was produced and this led to faster drying of the grains as higher values of flow were correspondingly displayed. The meter measures values accurately to over 1x102m/s. KEYWORDS: Airflow, digital meter, grain, drying. 1.

INTRODUCTION

When air is forced through a bulk crop, it must travel through narrow paths between individual particles (Williams, 1993). For packaged crops, air must travel through or between individual containers. Friction along air paths creates resistance to airflow. Fans must develop enough pressure to overcome this resistance and move air through the crop. The Agricultural Engineer is constantly exploring ways of obtaining data handling and processing, and the instruments available are often crude and prone to errors owing to their analogue format. More modern digital instruments are therefore, required and desirable. In-bin control systems usually measure the temperature and relative humidity of the ambient air and the temperature of the grain. Some controllers also measure the relative humidity of the air in the interstices of the grain mass. The proper location of sensors in the bin is critical. Control actions are based on the maximum temperature and equilibrium relative humidity of the grain, and thus the sensors should be located where these values are likely to occur, namely in the center of the bin under the loading spout. Multiple sensors increase the chance of detecting a hot spot; the choice of the number of sensors is an economic compromise. Modern in-bin controllers are equipped with microprocessors that allow the user to change the strategy of the control action. It is clear that all eight criteria cannot be minimized simultaneously. Thus, the manager of a grain depot has to decide which performance criterion should be minimized before a microprocessor is programmed (CIGR, 1999). Higher static pressures decrease fan output. As air enters the grain, it picks up some moisture, which cools the air slightly. As air moves through a deep grain mass, the air temperature is gradually lowered and relative humidity increased until the air approaches equilibrium with the grain. If the air reaches equilibrium with the grain, it passes through the remaining grain without any additional drying. If high relative humidity air enters dry grain, some moisture is removed from the air and enters the grain. This slightly dried air will begin to pick up moisture when it reaches wetter grain. Air in a 12 to 16 inch grain column does not reach equilibrium with the grain.


65


2.


2.1

Theory of Air Flow Rate Essential in Estimating Fan Power Requirements

Fans, especially axial fans are very essential in continuous flow drying whereby they influence the speed and uniformity of drying. According to Onwualu et al (2006) the efficiency of an axial fan can be estimated using η =Pin/Pout where, ---------------------------------------------------------------(i) η =efficiency (%) Pin = input power (watts); Pout =output power (watts) The flow rate of drying air is also essential in determining the power output required in the equation stated above. So v = Ϭh where, v = air flow rate, m3/s ------------------------------------(ii) h = pressure head, m of water Ϭ = density of air kg/m3 Wilcke and Morey (2002) in their work concluded that fans are usually described by the horsepower (hp) rating of the motor used to drive the impeller. It's helpful when selecting fans to estimate the power requirement first so you know where to start looking in the manufacturer's catalog. Furthermore fan motor size depends on the total airflow being delivered, the pressure developed, and the impeller's efficiency. Impeller efficiencies generally range from 40% to 65%. If we assume an average value of 60%, we can use the following formula to estimate the fan power requirement. According to Hunt (1977), the power required by such fans discussed here is given by P =vh/ce

-------------------------------------------------------------------(iii)

Where: p = power required, kW (HP) V= air flow, m3/min (ft3/min) h = static pressure head, mm(in.) of water e = static efficiency, decimal c = constant, 6000 (6350) 2.2

Design Considerations

The selection of materials for the design took cognizance of the following basic considerations: i) Material Conductivity:- a wooden box, a poor conductor of both heat and electricity was chosen because of its ability to insulate and prevent heat transfer across the walls of the box from the surrounding environment which could lead to the malfunctioning of the meter; ii) A good conductor such as a metal has the tendency to short circuit the naked components installed on the board and cause the malfunction of the device, so the preferred choice of the wooden box; iii) The size of the circuit board was designed to be almost the same with the inner dimensions of the box to enable it fit tightly to avoid wobbling; iv) A leather outer cover of the wooden box selected to preserve it and to add aesthetic value to the meter.


66


2.3

Hardware Design

The major component making the hardware is a rectangular box which serves as the housing and protective cover to the internal components. It consists of a box of size 172x85mm and inner dimension, 160x73mm. An opening was made on the top right hand corner of the box to accommodate the LCD projected out to interface with the user. The power control unit and the dynamo are equally projected and fixed adjacent to the LCD on the box. The power unit is made up of the on off switch and the call Button which calls the written program whenever the device is switched on. 2.4

Design of Software

The computer programming Language used was Assembly Language. The coded instructions were programmed into the micro controller component of the system. This program as written in assembly language and burnt into the micro controller is illustrated in a logic flow chart for the digital meter as shown in Figure 1. Start

Read ADC Output through Port

Display Equivalent Velocity on LCD

Any change in ADC Value?

Yes

No Fig 1: Flow chart of the function of the digital flow meter 2.5

Assembly Procedure

The components were assembled together in a circuit board (vero board) measuring 171x91mm as shown in Fig.2 using a simple soldering process in accordance to electronic circuit principles.


67


Faraday’s Law states that the electro motive force( EMF) induced in a circuit is equal to negative change of magnetic flux with time through the circuit. This can be stated mathematically as follows:

d

 E.dl   dt .......... .......... .......... .......... .......... .......... .......... .....(1) OR

 E.dl 

d B.ds.......... .......... .......... .......... .......... .......... ......( 2) dt s

Where S = any surface whose ring is the loop under consideration, B=magnetic flux It is observed that rate of change of magnetic flux with time through a circuit is equal to the curl of EMF produced

 E.dl l=  curlE.ds.......... .......... .......... .......... .......... .......... .......( 3) s

dB

 curlE.ds   dt ds.......... .......... .......... .......... .......... .......... ..( 4) s

Curl E = 

XE 

B .......... .......... .......... .......... .......... .......... .......... ..(5) t

 B .......... .......... .......... .......... .......... .......... .......... ....( 6) t

From this observation, when a magnetic flux passes through a coil, an electric current is produced which is proportional to the rate of change of the magnetic flux. In application of this principle a mini dynamo which comprises of a movable magnet and fixed coil is used as sensor. A blade was fixed on the movable magnet in such a way that air flow can freely turn the blade thereby turning the magnet which then produces a change in magnetic flux which is proportional to the air flow through the blade. In line with Faraday’s law, this change in magnetic flux produces electromotive force, EMF which is proportional to the change in the magnetic flux caused by the speed of air flow through the blade. The diagrammatic representation of the top view, dimension view, side view, isometric view and the circuit diagram of the airflow meter are shown below in Figures 2-6 respectively.


68


Fig 2: Top view of the air flow meter

Fig. 3: Dimension view of the flow meter


69


Fig. 4: Side view dimensions of the airflow meter

Fig.5 Isometric view of the air flow meter. The circuit board contains the arrangement of all the electronic components used in the design and development of the instrument. This arrangement is represented in figure 6.


70


Figure 6: Circuit Diagram of the portable digital Air flow meter Although the rate of air flow has been converted into electrical quantity by the sensor (Mini dynamo) but it cannot be used for display nor computation since it is still in its analogue form. To convert this analogue voltage form the sensor into binary which the micro controller understands, we need an analogue to digital converter (ADC 0804). ADC 0804 is a twenty pin integrated circuit(IC) which monitors a change in voltage at its input (pin 6) and displays the result in binary form(as demonstrated in table 1) which is equivalent to the voltage at the input. This output is taken from pin 11 to pin 18. These actions of the Analogue to Digital Converter is shown in Table1 below. Table 1: Analogue to Digital Conversion actions of the Decoded output Voltage of the Digital Air Flow Meter Analogue voltage Enable pins Select pins Outputs value(ohms) causing output (indicating the state (level or decoded output voltage) of conversion) Input (pins) G1 G2 G3 A B C Y0 Y1 Y2 Y3 Y4 Y5 Y6 Y7 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 1 1 0 0 0 0 1 0 1 0 0 0 0 0 0 2 1 0 0 0 1 0 0 0 1 0 0 0 0 0 3 1 0 0 0 1 1 0 0 0 1 0 0 0 0 4 1 0 0 1 0 0 0 0 0 0 1 0 0 0 5 1 0 0 1 0 1 0 0 0 0 0 1 0 0 6 1 0 0 1 1 0 0 0 0 0 0 0 1 0 7 1 0 0 1 1 1 0 0 0 0 0 0 0 1 These values from the ADC are in volts and can only be observed using a voltmeter. In order to display these values in a way everybody will read and understand (high level language), liquid crystal display (LCD) is used through the help of micro controller which computes the machine language (binary code) and displays it in a high level language. According to Mazidi (2006) in his book “Micro controller and embedded systems”, a micro controller is a logic IC which can do nothing unless when programmed to do so. Micro controller used here is a forty Nigerian Institution of Agricultural Engineers © www.niae.net

71


(40) Pin IC which comprise of four ports, port zero to port three. Each port is a group of eight (8) pin of the IC. From Pin 32 to pin 39 is the port zero, from pin 1 to pin 8 is port one, from pin 21 to pin 28 is port two, and from pin 10 to pin 17 is port three. The value from ADC is fed into the micro controller through port one. This means that the eight (8) pin output (pin11 to pin 18) of the ADC is fed into port one of the micro controller (Pin 1 to pin 8). This connection is as shown in the circuit diagram in Fig.6. As these values have entered the micro controller, the only way for one to appreciate what is happening is when it is displayed in an LCD which acts as a monitor in a computer system. This LCD (Liquid Crystal Display) is a sixteen pin display device which is used to display alphabet and numbers (alpha-numeric). It is connected through port two and three of the micro controller as shown in the circuit diagram. This ADC, micro controller and LCD all work within the voltage limit of 5v. Although the connections are rightly placed but the micro controller will not understand nor do anything without set of instructions called programs. This is in line with the statement of Hunt, (1977). The program is the software aspect of the project which includes calibrations, display unit, timing unit, memory unit etc. It is also with this set of codes that the micro controller understands the codes from the ADC and interprets it in the LCD for people to understand. This is actually the intelligent part of this research work. 2.6

Mini Dryer

This is a device consisting of electric fan and mesh-like tray, all embodied in a rectangular pan for drying of grains. Although it is not part of the air flow meter, it is necessary to include it so as to simulate the air stream situation in the plenum of deep-bed system where the meter would normally work. The fan is situated facing downward (900) at the top of mesh-like tray such that air produced by the fan can easily pass through the grains on top of the mesh-like tray. When the fan is plugged in a mains supply, the air produced pass through the wet grain and blow away the moisture content of the grains. The speed of the fan or the position of the mesh-like tray can be altered depending on how moist the grain is or how fast the drying exercise is to take place. When the speed of the fan is increased, more air will be produced and this will lead to faster drying of the grains. Also, when the mesh-like tray is moved closer to the fan, the grains will dry faster. The air flow can be determined using the device described above. The sensor of the device is placed in between the air flow, [that is in between the fan and the grains]. As this device is moved closer to the fan or speed of the fan increased, the value of the air flow rate (m/s) increase


72


2.7

Performance Test

The meter was evaluated using the improvised mini dryer described in section 5 above using some samples of cereal grain of maize (corn), sorghum and rice. The fan located at a position (90 0) at the above the mesh-like tray loaded with samples of the grain. Two loading depths of 20mm and 40mm of the grain were maintained throughout the experiment: the corresponding distances (of the meter from the fan) of 40mm and 90mm were adopted for taking the readings of the air speed. The fan was plugged into the mains supply which made the fan to rotate very fast and the air produced transmitted an air stream to the meter through the externally fixed blade of the dynamo. This process was repeated two times for each sample of crop according the chosen levels of height and depth stated above. Various readings of the velocity of air were recorded. The speed of the fan or the position of the mesh-like tray can be altered depending on how moist the grain is or how fast the drying exercise is to take place. 3.

PERFORMANCE TEST RESULTS

3.1

Air Flow

Grains were placed on a mesh like bowel in the dryer. This was done in such a way that the air flowing from across the grains could pass from top to bottom of the mesh like bowel and even pass out through the mesh. This was also done in such a way that the moisture content of the grains can be carried away in the flowing air. This flowing air volume was measured relative to the time spent in seconds. The mini dynamo in the device built to have a blade which enables the flowing air to turn when the measuring device is brought beneath the blowing fan in the dryer at an angle of 900. As this dynamo (sensor) rotates, it produces an electric current which is directly proportional to the rate of Air flow passing through the mass. This electric current produced is fed to an Analog to Digital convertor (ADC). The ADC is a twenty pin IC which is meant to receive voltage from a sensor and convert it into binary codes. The output of the ADC has eight pins which correspond to a point of the Micro Controller. The device under the above arrangement displayed some figures rising rapidly from 0 in the following progression :-1m/s, 2ms, 3m/s 4m/s……………………nm/s. It is worthy to note that the speed of the meter reading varied with its distance from the fan as well as with the size of the fan used at any giving instance. Hence the device is always subject to detect any rate of discharge incident on it. The results obtained are presented in table 2. Table2: Results of Performance Test of the Portable Air Flow Digital Meter Crop Moisture content Distance of digital Depth of grain Recorded Airflow of grain % wb meter from fan in tray (mm) on meter(m/s) (mm) Red corn I 13.84 40 20 56 60 40 37 Red corn II 16.07 40 20 56 60 40 36 White corn 14.01 40 20 57 Nigerian Institution of Agricultural Engineers © www.niae.net

73


Sorghum

13.84

Rice

13.37

60 40 60 40 60

40 20 40 20 40

36 57 37 56 36

The velocity of air recorded at the instrument distance of 40 and depth of 20 mm was 56m/s and this was common for red corn I and II, and rice; then 57m/s for white corn and sorghum. This is an insignificant variation of 1m/s which may have been likely caused by some slight fluctuation in the value of current generating the flow. On the other hand, a common air flow value of 36m/s was read for the samples of red corn II, white corn and rice at the fan distance of 60mm and drying depth of 40mm respectively; while red corn I and sorghum at the same levels recorded the air speed of 37m/s. This similarly, may be attributed to the same factor of fluctuating current. It was observed generally, that moving the digital meter away from the fan by 20mm results in a drop in its recorded value. This phenomenon practically implies that the air velocity generated diminishes with distance moved away. 3.2

Moisture Content Results

The moisture content determination was carried out only on two different samples of maize grains and not on other crops. The results obtained for the moisture from the two samples are presented as in Tables 3 and 4. Table 3: Results of Moisture contents Test for White corn Weight Of Weight Of Air Weight Of Weight Of Empty Crucible Dry Oven Dry Water 16.1 24.3 23.2 1.1 16.3 26.9 25.4 1.5 16.1 24.7 23.5 1.2 TOTAL AV MC FOR SAMPLE I = 41.51/3 = 13.84% Table 4: Results of MC test for red corn Weight Of Weight Of Air Weight Of Empty Dry Oven Dry Crucible 16.1 30.8 28.4 16.1 32.6 29.9 16.1 32.2 29.7

Mc Wb % 13.41 14.15 13.25 41.51

Weight Of Water

Mc Wb %

2.4 2.7 2.5 TOTAL

16.32 16.36 15.53 48.21

AV MC FOR SAMPLE II = 48.21/3 = 16.07% 4.

CONCLUSIONS

The velocity of the air movement and absolute pressure exerted by the airflow within the open space and inside the grain pile and between the mesh and grain surface interface denote an important advancement in local technology. This indeed has a correlation with the outputs of developed technologies in the developed economies. For each mass flow case, the velocity magnitude and the absolute pressure exerted by the airflow increased near each of the air inlets. Hence, the findings suggest that the nearer the meter to the source of drying air, the higher the impact of the stream of air flow on the meter. This also implies Nigerian Institution of Agricultural Engineers © www.niae.net

74


that in a deep-bed drying system, the layers closer to the air blowers are likely to dry faster than those at the middle and the extreme end. This adequately agrees with the views of (Chakraverty, 1988; CIGR, 1999 and Ojha and Michael, 2006). REFERENCES Chakraverty, A. 1988. Post Harvest Technology of Cereals, Pulses and Oil Seeds; Revised edition, Indian Institute of Technology, Kharapur; Oxford and ISH Publishing Co. PVT Ltd New Delhi Bombay Calcuta. Pp13 – 91. CIGR Handbook, 1999. Hand book of Agricultural Engineering , Vol4;Agro-Processing. Pp30-47. Hunt, D. R. 1977. Farm Power and Machinery Management 7th edition, Iowa State University press. Ames Iowa, USA, Pp179-187. Mazidi, J.G. 2004. Micro controller and Embedded Systems,4th edition, perentic hall, USA. Ojha, T.P. and A. M. Michael 2006. Principles of Agricultural Engineering Vol.1 6 th edition, Jain Brothers (New Delhi) 16/873, East Park Road, Karol Bagh. Pp651-617. Onwualu, A. P; Akubuo C. O. and I. E. Ahaneku (2006). Fundamentals of Engineering for Agriculture 1st edition, Immaculate publishers, Enugu, Enugu State. Pp205 – 207. Wilcke, W. F. and R. V. Moorey 2002. Selecting Fans and Determining Air flow for Drying , Cooling and Storage; Regents of the University of Minnesota. FO-05716-GO, Pp7 – 25. Williams, D. 1993. Estimating Airflow for In-Bin Grain Drying Systems, University of Missouri Extension. Pp1 – 6.


75


CHARACTERIZATION AND DISAGGREGATION OF DAILY RAINFALL DATA OF ONITSHA, ANAMBRA STATE, NIGERIA G. I. Ezenne, H. J. Ugwuozor and C. C. Mbajiorgu Department of Agric. & Bioresources Engineering, University of Nigeria, Nsukka E-mail: [email protected] ABSTRACT Characterization and disaggregation of rainfall data is very important for locations where the existing rainfall data is scanty. This study characterizes and disaggregates the rainfall data of Onitsha, Anambra state, Nigeria, using simple mathematical and statistical models; especially the WMO guideline. The disaggregation process was carried out over a time series below 6hours for as small as 30 minutes to 10 seconds. Analysis of the rainfall characteristics of Onitsha showed a maximum intensity of 146.5mm/day occurring in March. Highest probability of dryness of 0.98 was observed in January and December. Maximum mean daily rainfall of 14.81mm and 14.51mm occurred in September and July respectively. Disaggregated results obtained with WMO guideline was compared with that of USDA SCS Generalized Accumulated rainfall curves. Both the curves derived from the WMO model and three USDA SCS Generalized Accumulated rainfall curves (A, B and C) increased together in a positive correlation but the WMO model’s curve came closest to the curve B of the USDA, SCS model. This implies that the maximum intensity is usually reached in the middle of the storm duration in the tool. 1.

INTRODUCTION

Hydrological processes are usually studied in different time scales. The problem is often how to generate consistent time series both in higher-level time scale and lower-level, time scale. An approach to solve this problem is to model the process in the lower-level time scale only, and then aggregate to derive the process in the higher-level time scale. In some cases, the higher-level process may be the output of a specialised model or known from measurements (e.g., daily rainfall measurements); in such cases the aggregation approach cannot work, but rather disaggregation is needed. This kind of problem is commonly tackled by disaggregation models. Several such models have been developed since the 1970s and utilised in numerous hydrological applications, including, among others, simulation of reservoir systems, either for design or operation purposes, storm and flood simulations, and even enhancement of hydrological data records. Since many scenarios occur at a coarse time scale, there is a need to transform them into a finer scale. However, according to Koutsoyiannis et al. ( 2003), disaggregation is not identical to downscaling, as the latter aims at producing finer scale time series with the required statistics but that do not necessarily add up to any given coarse scale totals. The guideline proposed by Michel and Ojha (2003) for the break-down and distribution of 6-hour rainfall over a 6-hour period is a very remarkable document in the research. Available rainfall records for Onitsha are limited to daily time steps. Though rainfall data at shorter time steps are important for various purposes like modelling of erosion processes and flood hydrographs, they are hardly available in Nigeria. Hence, the objectives of this study are to characterize the rainfall data of Onitsha, disaggregate the rainfall data into fine-time scale and compare the obtained disaggregated results with the USDA SCS generalized accumulated rainfall curves results. 2.

METHODOLOGY

2.1

Stochastic Rainfall Disaggregation: Review of Earlier Work

Stochastic disaggregation of rainfall is based on stochastic point processes. A simple stochastic point process of rainfall can be conceptualised by storms arriving in a Poisson process, and each storm is associated with a random and single rain cell of rectangular pulse with independent intensity and duration. Nigerian Institution of Agricultural Engineers © www.niae.net

76


The rain cells can overlap, and the total rainfall intensity at any instance can be given by the sum of the intensities of all active rain cells at that instance. One of the shortcomings of this type of conceptual model is that they are not capable of accounting for the temporal statistics of rainfall at different aggregation levels (Rodriguez-Iturbe et al., 1987). Cluster based models such as Neyman-Scott and Bartlett-Lewis were introduced by Rodriguez-Iturbe et al., (1987 and 1988) as measures to overcome the shortcomings of the point processes. These cluster based rainfall models are actually based on the Poisson process with some adjustments. In the cluster based models, each storm produces a cluster of rain cells instead of one as in the simple Poisson point process, with each cell having random duration and intensity. From a statistical point of view both randomised and nonrandomised versions of the models are parametric analyses. Normally, these parameters are estimated by the method of moments. Some of the remarkable works on stochastic rainfall disaggregation include the works by Cowpertwait et al. (1996) on disaggregation of hourly rainfall into smaller time intervals by allocating pulses of a specified small depth each, at the different intervals and Connolly et al. (1998) disaggregation of daily rainfall into a number of events following a Poisson process. A more systematic model, based on the Bartlett-Lewis rectangular pulse process, was studied by Koutsoyiannis et al., (2001). The abovementioned models are generally based on classical probability and stochastic processes theory. In the last years there have been studied disaggregation techniques with different scientific bases. Thus, the development of multifractal simulation techniques has provided a potentially powerful tool for the exploration of problems such as disaggregation. An application of this approach to the disaggregation problem was proposed by Olsson and Berndtsson (1997). The use of a self-similar micro canonical cascade enables the reproduction of the exact total daily rainfall, but it does not allow for the reproduction of the observed hourly autocorrelations. Bounded micro canonical cascades (Marshak et al., 1994) do however provide a tool, which could be used for disaggregation. Such approaches are promising, as illustrated by the successful reproduction of rainfall statistics with canonical bounded cascades (Menabde et al., 1997), but require more analysis, particularly in their ability to reproduce the dry period structure at different scales. A different approach has been followed by Burian et al., (2000) who used artificial neural networks to disaggregate hourly rainfall data into shorter time intervals. Yet another approach was followed by Sivakumar et al. (2001) who formed a simple chaotic model to disaggregate rainfall of five resolutions using techniques from the chaos literature. It must be mentioned however that the latter techniques that are not based on probability and stochastic processes may not be appropriate for large length simulations, as they do not perform well in extrapolation. For example they may result in poor description of extremes, whose study needs large length simulations. A successful rainfall generation model capable of reproducing the particular fine scale structure of rainfall can be used as the lower-level model. This can then be combined with an appropriate procedure for adjusting the lower-level amounts so as to obtain the required higher-level totals. Such an implementation was done by Koutsoyiannis et al., (2001) and resulted in a computer program named Hyetos that can be easily applied at any location provided that a minimal amount of data is available that can support parameter estimation. The higher- and lower-level time scales in this implementation are respectively daily and hourly which are the most suitable for typical hydrological applications. 2.2

The Study Area

Onitsha is located in the south-eastern part of Nigeria. The area experiences highest daily rainfall in the months of July through September. There are two seasons in Onitsha similar to every other part in Nigeria, the wet season (April-October) and dry season (November-March).The dry season starts with Harmattan - a dry chilly spell with a dusty atmosphere brought about by the NE winds blowing from the Arabian Peninsula across the Desert. During the rainy season, a marked interruption in the rains occurs during August, resulting in a short dry season often referred to as the “August break”, though for years now this has not been consistent in August due to climate change. In the area, humidity and temperature Nigerian Institution of Agricultural Engineers © www.niae.net

77


are relatively high year-round. Though, humidity is lower in December and January during the Harmattan or dry season when cool dry winds blow off the desert (Okonkwo and Mbajiorgu, 2010). 2.3

Disaggregating Daily Rainfall into Individual Depths

The relationship between the small rainfall amount and the daily total amount is given as follows: Nc

y

ci

 Zc

…………………………………………………….. (1) Where Nc = Number of Yci amounts of rainfall, Zc = daily total amount of rainfall, Yci = small rainfall amounts. i 1

Transformation of the daily rainfall amount was done using the transformation model proposed by Hershenhorn and Woolhiser (1987). This transformation was necessary to convert the observed rainfall data to a value that will be compatible with a typical disaggregation model. The transformation is given as follows: 

 Z ……………………………………………………. (2)   

F (Zc) =1-exp 

where  = 7.617, γ =0.7364, Z = daily total amount of rainfall and F (Zc) = transformed daily total amount of rainfall Usually, if there is scanty information about the actual starting time of a rainfall event, duration and interval, the local starting time is assumed. In the case of Onitsha meteorological station, there is no record of starting time, hence the need to assume them. This assumed local starting time also has to be transformed especially if the results will be applied to a disaggregation model. The transformation is done using the following relationship equation: Tc =

Uci ……………………………………………………. (3) 24

where Uci = observed starting time, Tc = transformed starting time The above result is a general approach to the generation of a typical raw data for observed daily rainfall. It does not specify for an exact environmental purpose. But for agricultural purposes and watershed management in agriculture the time of concentration is often 6-hours (Michel and Ojha, 2003). 2.4

World Meteorological Organization (WMO) Guideline for Breaking Down 6 Hours Rainfall Data

The WMO guideline is a simple linear distribution and conformed to the simple linear regression model given by; Y =  + x +  ………………………………………………… (4) Where Y = percentage of rain, X = duration of rain, regression parameters.

 = superimposed error component,  and  =

 yi   xi   Y  X n But  = n ……………………………………… (5)


78


where n = number of observed data



 xiyi   xi  yi …………………………………………………. (6) 2 n  xi2   xi

where Ῡ= mean percentage of rain, X = mean duration of rain, Yi= ith percentage of rain and Xi= duration of ith rain. Table 1 shows the estimation of the WMO model parameters. Table 1: Estimation of the WMO model parameters S/N X Y X2 1 0 0.0 0 2 0.5 2.0 0.25 3 1.0 8.0 1.00 4 1.5 15.0 2.25 5 2.0 22.0 4.00 6 2.5 60.0 6.25 7 3.0 70.0 9.00 8 3.5 78.0 12.25 9 4.0 84.0 16.00 10 4.5 88.0 20.25 11 5.0 92.0 25.00 12 5.5 96.0 30.25 13 6.0 100.0 36.00 39.0 715.0 162.5 

XY 0 1.0 8.0 22.5 44.0 150.0 210.0 273.0 336.0 396.0 460.0 528.0 600.0 3028.5

 3.3 -4.4 -8.1 -10.8 -13.5 14.8 15.1 13.4 9.7 4.0 -17.0 -7.4 -13.1 -13.8

The error in given by: I = Yi -  - xi ……………………………………………………. (7) Using the above equations and parameter values, the relationship is developed as follows: Yi =- 3.3 + 19.4 Xi + I …………………………………………… (8) The extension is between 0.00 and 0.50 on X- axis and the error is determined by interpolation.

 f   i  f x



(Yi   i ) F  (Yi   i )i (Yi   i ) F  (Yi   i) x

……………………………… (9)

where x= the unknown error. 2.4

Disaggregated and Time Distributed Rainfall Data

According to Hershenhorn and Woolhiser (1987) the joint distribution function of the number of storm events and the transformed daily amount provided resolutions for one storm event and as much as six storm events on a rain day. However it was observed that there can only be a maximum of four storm events and a minimum of one storm event on a typical rain day. Hence both the assumed number of storm events and the individual storm depths were made to satisfy the conditions of aggregation and observable number of events. Applying the WMO model with the observed duration of individual rainfall depths, the time distributed rainfall depths were gotten and compared with the assumed individual storm depths.


79


Both the starting time and duration of individual storms were assumed, however the time distributed rainfall depth was arrived at by multiplying the percentage ratio of rainfall amount corresponding to the duration of individual storm events as suggested by the WMO guideline, with the converted 6-hour rainfall of a given daily rainfall total. This will generate results of time distributed and disaggregated rainfall data which sums up as closely as possible to the coarse or 24-hour rainfall total. 2.

RESULTS

3.1

Daily Rainfall Characterization Results

The daily rainfall characteristics of Onitsha are summarised in Table 2 below. It was obtained by statistically analysing the daily rainfall record from the month of January to December for 30 years. Table 2: Daily Rainfall Characteristics of Onitsha Months Mean STD Var. Prob Rainfall (dry) intensity January 10.35 14.27 203.71 0.98 February 11.46 12.81 164.06 0.97 March 11.88 18.42 339.38 0.90 April 14.03 15.68 245.93 0.87 May 15.71 18.17 330.01 0.56 June 13.84 17.48 305.54 0.37 July 14.51 19.42 376.95 0.05 August 11.69 17.64 311.11 0.07 September 14.81 19.14 366.38 0.20 October 13.19 17.21 296.38 0.68 November 6.98 9.36 87.67 0.89 December 7.18 7.90 62.45 0.98

C.V.

0.73 0.89 0.64 0.89 0.86 0.79 0.75 0.66 0.77 0.77 0.75 0.91

Mean rain days 0.84 2.11 4.71 9.23 14.16 18.00 19.58 21.13 20.53 19.19 3.83 0.97

Max. intensity 59.2 46.0 146.5 102.8 98.3 97.1 172.4 137.0 94.5 124.8 55.6 28.1

No of wet/dry spell days 1/30 1/27 1/21 2/14 2/5 3/2 7/1 6/1 3/2 2/5 1/15 1/30

Observed and Transformed Rainfall Data: Table 3 below describes the observed and transformed rainfall data. It shows the transformed results computed as explained previously. Table 3: Observed and Transformed Rainfall Data S/ No. of Observe Transfor Observed N Rainfa d coarse med fine-scale ll rainfall coarse rainfall events depth rainfall depth (min) depth (mm) (mm) 1 2 10.35 0.7144 5.10 5.25 2 3 11.46 0.7410 3.05 3.95 4.46 3 2 11.88 0.7502 5.90 5.98 4 3 14.03 0.7915 4.50 4.95 4.58

Transfor med finescales rainfall depth (mm) 0.5250 0.5325 0.3993 0.4602 0.4905 0.5633 0.5669 0.4927 0.5172 0.4972


Observ ed starting time (hr)

Transfor med start time (hr)

Observ ed rainfall duratio n (Min)

Transfor med rainfall duration (Min)

00.30 01.30 02.30 03.30 04.30 05.30 06.30 07.30 08.30 09.30

0.0125 0.0542 0.0958 0.1375 0.1792 0.2208 0.2625 0.3042 0.3458 0.3875

3.50 4.50 2.50 1.50 5.00 3.00 5.50 3.50 3.50 1.00

1.253 1.504 0.916 0.405 1.609 1.099 1.705 1.253 1.253 0.000 80


5

4

15.71

0.8181

6

2

13.84

0.7882

7

4

14.51

0.7996

8

3

11.69

0.7461

9

4

14.81

0.8044

10

3

13.19

0.7765

11

3

6.98

0.6085

12

2

7.18

0.6161

3.90 4.10 3.50 4.21 6.00 7.84 3.65 3.00 4.85 3.01 4.00 3.85 3.84 3.70 3.00 4.95 3.16 4.30 4.95 3.94 2.30 3.55 1.13 3.90 3.28

0.4571 0.4694 0.4311 0.4760 0.5678 0.6399 0.4411 0.3956 0.5119 0.3963 0.4633 0.4540 0.4533 0.4443 0.3956 0.5172 0.4074 0.4813 0.5172 0.4596 0.3390 0.4345 0.2176 0.4571 0.4159

10.30 11.30 12.30 13.30 14.30 15.30 16.30 17.30 18.30 19.30 20.30 21.30 22.30 23.30 00.30 01.30 02.30 03.30 04.30 05.30 06.30 07.30 08.30 09.30 10.30

0.4290 0.4708 0.5125 0.5542 0.5958 0.6375 0.6792 0.7208 0.7625 0.8042 0.8458 0.8875 0.9292 0.9708 0.0125 0.0542 0.0958 0.1375 0.1792 0.2208 0.2625 0.3042 0.3458 0.3875 0.4292

3.50 2.50 1.00 1.50 3.50 4.50 2.00 2.00 2.50 2.50 2.00 2.50 3.50 1.50 2.00 2.50 3.00 2.00 3.50 2.50 3.00 3.50 1.50 4.00 3.50

1.253 0.916 0.000 0.405 1.253 1.504 0.693 0.693 0.916 0.916 0.693 0.916 1.253 0.405 0.693 0.916 1.099 0.693 1.253 0.916 1.099 1.253 0.405 1.386 1.253

Statistical Extension of the WMO Guideline: The WMO guideline provided percentage rainfall resolutions for period of rainfall at a 0.5 interval. But the resolution is required in periods below 0.5hours. Table 4 shows the statistical extension of the WMO Model Table 4: Statistical Extension of the WMO Model S/N Time (hr) or x % of rain or y 1 0.000 0.000 2 0.001 0.004 3 0.005 0.021 4 0.010 0.039 5 0.015 0.060 6 0.020 0.080 7 0.050 0.200 8 0.100 0.400 9 0.150 0.600 10 0.200 0.800 11 0.250 1.000 12 0.300 1.200 13 0.350 1.400 14 0.400 1.600 15 0.450 1.800 16 0.500 2.000


(y-) -3.300 -3.280 -3.200 -3.110 -3.010 -2.910 -2.330 -1.360 -0.390 0.580 1.550 2.520 3.490 4.460 5.430 6.400

Error () 3.300 3.284 3.221 3.149 3.070 2.990 2.530 1.760 0.990 0.220 -0.550 -1.320 -2.090 -2.860 -3.630 -4.400

81


17 18 19 20 21 22 23 24 25 26 27

1.000 1.500 2.000 2.500 3.000 3.500 4.000 4.500 5.000 5.500 6.000

8.000 15.000 22.000 60.000 70.000 78.000 84.000 88.000 92.000 96.000 100.000

16.100 25.800 35.500 45.200 54.900 64.600 74.300 84.000 109.000 103.400 113.100

-8.100 -10.800 -13.500 14.800 15.100 13.400 9.700 4.000 -17.000 -7.400 -13.100

Disaggregated and Time Distributed Data Results: Table 5 below gives a general overview of the disaggregated rainfall data over different time intervals. For each event of a given rainfall duration, there is a corresponding disaggregated result to be expected. Table 5: Disaggregated and Time Distributed Data S/N No of Coarse Total Disaggregated events Rainfall daily Rainfall depth Depth rainfall 1

2

6.210

10.35

2

3

6.878

11.46

3

2

7.128

11.88

4

3

8.418

14.03

5

4

9.426

15.71

6

2

8.304

7

4

8.706

8

3

7.014

11.69

9

4

8.886

14.81

10

3

7.914

13.19

Yc1 =5.10 Yc2 = 5.25 Yc3 = 3.05 Yc4 = 3.95 Yc1 =4.46 Yc2 =5.90 Yc3 = 5.98 Yc4 = 4.50 Yc1 = 4.95 Yc2 =4.58 Yc3 = 3.90 Yc4 =4.10 Yc1 = 3.50 Yc2 = 4.21 Yc3 = 6.00 Yc4 = 7.84 Yc1 = 3.65 Yc2 =3.00 Yc3 = 4.85 Yc4 = 3.07 Yc1 = 4.00 Yc2 =3.85 Yc3 =3.84 Yc4 = 3.70 Yc1 = 3.00 Yc2 = 4.95 Yc3 = 3.16 Yc1 =4.30


Duration of Local rainfall (hr) Starting time 3.50 4.50 2.50 1.50 5.00 3.00 5.50 3.50 3.50 1.00 3.50 2.50 1.00 1.50 3.50 4.50 2.00 2.00 2.50 2.50 2.00 2.50 3.50 1.50 2.00 2.50 3.00 2.00

00.30 01.30 02.30 03.30 0.4.30 05.30 06.30 07.30 08.30 09.30 10.30 11.30 12.30 13.30 14.30 15.30 16.30 17.30 18.30 19.30 2.30 21.30 22.30 23.30 00.30 01.30 02.30 03.30

Time distributed and disaggregated rainfall 4.844 5.465 4.127 1.032 6.328 4.990 6.843 6.566 6.566 0.168 7.352 5.656 0.784 1.414 6.477 7.308 1.915 1.915 5.224 5.224 1.543 4.208 5.471 1.333 1.955 5.332 6.220 1.741 82


11

3

4.188

6.98

12

3

4.308

7.18

Yc2= 4.95 Yc3 =3.94 Yc1 =2.30 Yc2 =3.55 Yc3 =1.13 Yc1 =3.90 Yc2 =3.28

3.50 2.50 3.00 3.50 1.50 4.00 3.50

04.30 05.30 06.30 07.30 08.30 09.30 10.30

6.173 4.748 2.932 3.267 0.628 3.619 3.360

30 Minutes Rainfall Data of Onitsha: Table 6 below shows the results of 30 minutes rainfall expected using the WMO guideline, for the rainfall record of Onitsha. Table 6: Thirty Minutes Rainfall Data of Onitsha S/N No of Total 24-hr. (mm) Local rainfall daily disaggregated starting events rainfall rainfall time(hr)

13.84

5.10 5.25 3.05 3.95 4.46 5.90 5.98 4.50 4.95 4.58 3.90 4.10 3.50 4.21 6.00

00.30 01.30 02.30 03.30 04.30 05.30 06.30 07.30 08.30 09.30 10.30 11.30 12.30 13.30 14.30

30-mins disaggregated rainfall result (mm) 0.0612 0.0630 0.0366 0.0474 0.0536 0.0708 0.0718 0.0540 0.0594 0.0550 0.0468 0.0492 0.0420 0.0506 0.0720

14.51

7.84 3.65

15.30 16.30

0.0940 0.0438

4.70 2.19

11.69

3.00 4.85 3.01 4.00

17.30 18.30 19.30 20.30

0.0360 0.0582 0.0362 0.0480

1.80 2.91 1.81 2.40

3.85 3.84 3.70 3.00 4.95 3.16 4.30 4.95 3.94 2.30

21.30 22.30 23.30 00.30 01.30 02.30 03.30 04.30 05.30 06.30

0.0462 0.0460 0.0444 0.0360 0.0594 0.0380 0.0516 0.0594 0.0472 0.0276

2.31 2.30 2.22 1.80 2.97 1.90 2.58 2.97 2.36 1.38

1

2

10.35

2

3

11.46

3

2

11.88

4

3

14.03

5

4

15.71

6

2

7

4

8

3

9

4

14.81

10

3

13.19

11

2

06.98


6hrs disaggregated rainfall results 3.06 3.15 1.83 2.37 2.68 3.54 3.59 2.70 2.97 2.75 2.34 2.46 2.10 2.53 3.60

83


12

2

07.18

3.55 1.13 3.90 3.28

07.30 08.30 09.30 10.30

0.0426 0.0136 0.0468 0.0394

2.13 0.68 2.34 1.97

Fine-time Scale Rainfall Data Results: Other smaller resolutions for the disaggregated rainfall results are presented in Table 7 below. This is done by applying the WMO guideline. Table 7: Fine Time Scale Rainfall Data S/N 6hrs 15 min disaggregated disaggregated rainfall depth rainfall depth (mm) (mm) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

3.06 3.15 1.83 2.37 2.68 3.54 3.59 2.70 2.75 2.34 2.46 2.10 2.53 3.60 4.70 2.19 1.80 2.91 2.40 2.31 2.30 2.22 1.80 2.97 1.90 2.58 2.97

0.0306 0.0315 0.0183 0.0237 0.0268 0.0354 0.0359 0.0270 0.0275 0.0234 0.0246 0.0210 0.0253 0.0360 0.0470 0.0219 0.0180 0.0291 0.0240 0.0231 0.0230 0.0222 0.0180 0.0297 0.0190 0.0258 0.0297

9 min disaggregated rainfall depth (mm) 0.0184 0.0189 0.0110 0.0142 0.0161 0.0212 0.0215 0.0162 0.0165 0.0140 0.0148 0.0126 0.0152 0.0216 0.0282 0.0131 0.0108 0.0175 0.0144 0.0139 0.0138 0.0133 0.0108 0.0178 0.0114 0.0155 0.0178


60 sec disaggregated rainfall depth (mm) (x 104) 20.196 20.790 12.078 15.642 17.688 23.364 23.694 17.820 18.150 15.444 16.236 13.860 16.698 23.760 31.020 14.454 11.880 19.206 11.946 15.840 15.246 15.180 14.652 11.880 19.602 12.540 17.028



84


Comparison of 2hr Disaggregated Rainfall with the USDA Generalized Accumulated Rainfall Curves The USDA SCS (1955) generalized accumulated rainfall curve gives the breakdown of percentages of rainfall that can be expected for a proportionate duration of storm. It comprises of three curves (Figure 1); curve A is for advanced type in which the highest intensity occurs in the early part of the storm, Curve B is for the highest intensity occurring in the middle of the storm, and curve C is for the highest intensity occurring late in the storm duration. Table 8 shows the comparison of the results generated from the disaggregation of Onitsha as compared to the results of curve A, curve B and curve C of the USDA SCS (1955) model

Fig 1: Generalized accumulated rainfall curves for A (advanced), B (intermediate), and C (retarded) types of storms. (USDA SCS, 1955) Table 8: Comparison of Results S/N Disaggregated Rainfall Result 1 0.6732 2 0.6930 3 0.4026 4 0.5214 5 0.5896 6 0.7788 7 0.78998 8 0.5940 9 0.6534 10 0.6050 11 0.5148 12 0.5412

Curve A result (mm) 1.9278 1.9845 1.1529 1.4931 1.6884 2.2302 2.2617 1.7010 1.8711 1.7325 1.4742 1.5498


Curve B result (mm) 0.6732 0.6930 0.4026 0.5214 0.5896 0.7788 0.7898 0.5940 0.6534 0.6050 0.5148 0.5412

Curve C result (mm) 0.5202 0.5355 0.3111 0.4029 0.4556 0.6018 0.6103 0.4490 0.5090 0.4675 0.3978 0.4182

85


3.2

Daily Rainfall Data Characterization and Data Analysis

The average number of rain day varied from 0.84 days in the dry month to 21 days in the wet month. The wet period is greater than the dry period. The wet/dry spell length varied from 1/30 in the dry season to 7/1 in the rainy season. This indicated the occurrence of wet and dry spell days in groups which is caused by the persistence of synoptic scale weather system. Temporal Characteristics of Rainfall Variation The variability of the daily rainfall intensity was high in all months as depicted by larger coefficients of variation. The mean daily rainfall intensity varied from about 6mm in dry seasons to 15mm in the wet season. Overall, the daily rainfall intensity was highest in the middle of storm periods. The observed maximum intensity varied from 46mm to about 170mm, while the dry probability varied from 0.91 in dry seasons to 0.66 in wet seasons. Variation of Coarse and Disaggregated Data The coarse rainfall data is the 24-hours rainfall data while the fine or disaggregated rainfall data is the rainfall data below 6-hour. The individual depths were resolved so that they sum up to the coarse depth as close as possible. This is reflected in the positive correlation of the coarse and disaggregated data. Both increased and decreased together. Comparison with the Generalized Accumulated rainfall curves The disaggregated rainfall results of Onitsha was applied to the generalized accumulated curve at various durations below 6-hour. Comparing the result with the ratio of each curve (figure 2) showed a close correlation of the rainfall result of Onitsha with curve B, this implies that the maximum intensity is usually reached in the middle of the storm duration.

RATIO OF RAIN 1.2

1

Curve C

0.8

Series1 Series2 Series3 Series4

WMO Curve

0.6

Curve A

0.4

Curve B

Curve C

0.2

0 0

1

2

3

4

5

6

7

TIME Figure 2: Comparison with the Generalized Accumulated rainfall curves (USDA 1955)


86


3.

CONCLUSIONS

The simple regression model was a good fit for the WMO guideline. This model helped in the statistical extension of the guideline so as to obtain resolutions for smaller percentage of rainfall at smaller time intervals. However more complex statistical models and mathematical approach should provide more precise results for fine-scale rainfall resolutions. The rainfall characteristics of Onitsha can be summarized as predominantly rainy, having a total of seven wet months and five dry months i.e. April to November and November to March. The daily rainfall characteristics were similar to tropical climates. Most of the rainfall occurred in the morning and evening hours. Also comparing the result with the ratio of each curve of the generalized accumulated rainfall curves (USDA SCS 1955) showed a close correlation of the rainfall result of Onitsha with curve B. This implies that the maximum intensity is usually reached in the middle of the storm duration. REFERENCES Burian, S.J., Durrans, S.R., Tomic, S., Pimmel, R.L., Wai, C.N. 2000. Rainfall disaggregation using artificial neural networks. J. Hydrol. Engng ASCE 5, Pp 299–307 Cowpertwait, P.S.P., O’Connell, P.E., Metcalfe, A.V., Mawdsley, J.A. 1996. Stochastic point processes modelling of rainfall, II, Regionalisation and disaggregation. J. Hydrol.,Vol. 175, Pp 67-65. Connolly, R. D., Schirmer, J., Dunn, P. K. 1998. A daily rainfall disaggregation model.Agricultural And Forest Meteorology, Vol. 92 (2), Pp. 105-117 Hershenhorn, J., and D.A. Woolhiser, 1987. Disaggregation of daily rainfall. J. of Hydrol.,Vol 95, Pp299322. Koutsoyiannis, D., C. Onof, and H. S.Wheater 2001. Stochastic disaggregation of spatialtemporal rainfall with limited data, 26th General Assembly of the European Geophysical Society, Geophysical Research Abstracts, Vol. 3, Nice, March 2001, European Geophysical Society. Koutsoyiannis, D., C., Onof, and H.S. Wheater, 2003. Multivariate rainfall disaggregation at a fine time scale, Water Resources Research (in press). Marshak, A., Davis, A., Cahalan, R. and Wiscombe, W. 1994. Bounded cascade models as nonstationary multifractals. Phys. Rev. E, vol. 49(1), pp. 55-69. Menabde, M., Harris, D., Seed, A., Austin, G. and D. Stow, 1997. Multiscaling properties of rainfall and bounded random cascades. Water Resour. Res., Vol. 33(12), Pp. 2823-2830. Michel A. M and T. P Ojha 2003. “Principle of Agricultural Engineering” 3rd edition. Jain Brothers publishers New Delhi, India.Vol II Okonkwo, G.I. and C.C. Mbajorgu 2010. “Rainfall intensity duration- frequency analysis for south eastern Nigeria”. Agricultural Engineering International : the CIGR Ejournal, Manuscript 1304. Vol. XII, Pp. 17 Olsson, J. and R. Berndtsson 1997. Temporal rainfall disaggregation based on scaling properties. Third International Workshop on Rainfall in Urban Areas, IHP-V, Technical Documents in Hydrology, Rodriguez-Iturbe, I., Cox, D.R., Isham, V. 1987. Some models for rainfall based on stochastic point processes. Proc. R. Soc. Lond.Vol 410, pp269-298. Rodriguez-Iturbe, I., Cox, D.R., Isham, V. 1988. A point process model for rainfall: Further developments, Proc. R. Soc. Lond., A Vol 417,pp 283-298. Sivakumar, B., Sorooshian, S., Gupta, V.J., Gao, X. 2001. A chaotic approach to rainfall disaggregation. Water Resour. Res., vol. 37, pp. 61–72. USDA SCS 1955. Soil and Water Conservation Engineering. Central Technology Unit. P. 20.


87


COMPARISON OF ACRU AND HEC-HMS MODELS IN RUNOFF PREDICTION IN A WATERSHED, SOUTHWEST NIGERIA A. P. Adegede1, K. N. Ogbu2, V. Ogwo3, C. C. Mbajiorgu3 1 Raw Materials Research & Development Council, Lokoja, Nigeria E-mail: [email protected] 2 Agricultural and Bioresources Engineering Department, Nnamdi Azikiwe University, Awka, Nigeria 3 ESHRU, Agricultural and Bioresources Engineering Department University of Nigeria, Nsukka, Nigeria ABSTRACT The frequency in the occurrence of hydrological extremes has necessitated the use of models to forecast and mitigate such disasters. Agricultural Catchments Research Unit (ACRU) model and Environmental Hydrology, University of Natal and the Hydrologic Engineering Centre’s Hydrologic Modelling System (HEC-HMS) were used to simulate runoff from Agbogbo catchment in Ile-Ife (Southwestern Nigeria). The objective of the study was to compare both models performance in simulating streamflow in the Agbogbo catchment. Model input data for both models were obtained and used for model simulation. Three years streamflow records were obtained for Agbogbo catchment and used for calibrating and validating the models. Data for the first hydrologic year (April 1987–March 1988) was used to calibrate the two models while that of the two remaining years was used for validating the models. When the NashSutcliffe Model Efficiency Coefficient was used to evaluate the performance of the models, a value of 0.83695 and 0.78232 was obtained for the ACRU model simulation for the 2nd and 3rd hydrologic years while values of 0.87748 and 0.54706 were obtained by the HEC-HMS model for the two periods. ACRU model performed better with an overall (April 1988–March 1990) model efficiency coefficient of 0.81218 when compared with that of HEC-HMS which was obtained to be 0.71498. When the hydrographs of the observed and simulated streamflow were compared, ACRU model showed more sensitivity to changes within the catchment as demonstrated by the similarity in the timing of the peaks than HEC-HMS, which implies that ACRU model can be used for flood forecasting in Agbogbo catchment. KEYWORDS: Digital elevation model, agrohydrology, unguaged catchments. 1.

INTRODUCTION

Continued land development and land-use changes within cities and at the urban fringe present considerable challenges for environmental management (Muthukrishnan et al, 2006). With increasing population and industry, the demand for water has increased prodigiously thereby imposing a higher efficiency in the planning and management of water resources. With streamflow accounting for only 0.006% of freshwater resources (Gleick, 1996), realistic and accurate streamflow forecasts have become an essential tool for water resources planning and management (Hobson, 1997). The prospect of adverse climate change is not going to diminish in the near future (Downing et al, 1997). Climate change could alter the timing, magnitude and duration of rainfall and other weather events. All evidence shows that climate variability has increased to such a degree that predictability of water availability has been reduced dramatically: weather extremes are shifting and intensifying, and thereby introducing greater uncertainty in the quantity and quality of our water supplies over the short and the long term (United Nations, 2009). Hundreds of rainfall-runoff models have been developed throughout the world, especially in Europe to provide river flow forecasting (Beven, 2001). Hydrological models have been developed to improve our understanding of surface runoff generated from complex watersheds, make efficient and cost effective quantitative estimates of water resources of unguaged catchments, and to plan, design, operate and manage water related structures.


88


Our ability to predict the hydrology of streams in future climates depends in part on our ability to model present circumstances. The comparison of observed to modelled streamflow provides insight into model performance and the ability to predict hydrologic attributes that might be of interest in future scenarios and extreme events (Whitfield et al, 2003). To gauge how well simulations perform requires rigorous assessment, and setting benchmark against which to measure success. Model validation is essential to the interpretation of simulation results. It illuminates under what circumstances a model reproduces events accurately and under what circumstances it performs unsatisfactorily. Validation is also critical to the improvement of models; the modelling community cannot improve models if it does not know how, where, and when they fail (Gordon et al, 2004). The objective of the study was to compare performance of the HEC-HMS and ACRU models in simulating stream flow in Agbogbo catchment, south eastern Nigeria. ACRU model (Schulze, 1989) was used in this comparison because it was developed in Africa and has been validated for several catchments in many parts of Africa. As a result of the dendritic drainage pattern of Agbogbo catchment, HEC-HMS (Scharffenberg and Fleming, 2010) which is a model developed for that kind of drainage geomorphology was also selected for this study. 2.


2.1

ACRU Model

Agricultural Catchments Research Unit (ACRU) model (Schulze, 1989) was developed by the former Department of Agricultural Engineering, now School of Bioresources Engineering and Environmental Hydrology, of the University of Natal in South Africa. It is a physical-conceptual rainfall-runoff model that simulates stormflows and baseflows explicitly, with a modification enabling the simulation of through flow (New, 2002). The ACRU model is a multi-purpose and multi-level integrated physical conceptual model that can simulate streamflow, total evaporation, and land cover/management and abstraction impacts on water resources at a daily time step. ACRU is highly versatile with potential applications ranging from streamflow simulation, to crop yield estimations, irrigation estimations, risk analysis etc. It has been mostly applied in the temperate and humid parts of South Africa and has been frequently used for assessing the impacts of various land use modifications, specifically commercial afforestation (Hugh, 2002).


89


The ACRU modelling system is made up of a number of discrete, but interlinked components. The linkages and components are illustrated in Fig 1.

OBSERVED DATA

ACRU INPUT UTILITIES AND DECISION SUPPORT SYSTEM

Daily hydrometerological data file (single or composite format)

Static input information menu

Output variables to be stored

Dynamic input information file (optional)

ACRU MODEL KEY FILE PROGRAM

Output Files

ACRU OUTPUT UTILITIES

Fig. 1: Components and Linkages of ACRU Modelling System (Smithers et al, 1994) The minimum daily input requirements are precipitation and potential evaporation. Parameter values for evapotranspiration, soil moisture budgeting and runoff generation are also required. ACRU simulates soil moisture in a vertical, two-layer soil column. Incoming rainfall is subject to interception by vegetation depression storage. The remaining rainfall infiltrates the upper soil horizon, and subsequently, moisture in excess of drained upper limit (that is field capacity) drains to the subsoil horizon. Similarly excess water in the subsoil horizon drains, either laterally as throughflow to the stream channel, or vertically to a groundwater store. Evapotranspiration occurs from both the topsoil and subsoil horizon, and is a function of potential evaporation (A-pan), leave area and soil moisture availability. When soil moisture is not a limitation, evapotranspiration occurs at the potential rate, but decrease linearly with increasing water stress once a critical fraction of plant-available water is reached. Surface runoff and infiltration are simulated using a modified form of the SCS equation (Schmidt and Schulze, 1987), viz. Q = (Pn – cS)2/[Pg + S(1 – c)]

1

Q is the runoff depth; Pn is the net daily rainfall (i.e. gross rainfall Pg, less canopy interception, plus contributions from impervious areas); S is the potential maximum retention (a function of soil texture and antecedent soil moisture); and c is the coefficient of initial abstraction. ACRU employs the continuity equation in routing flow through reservoirs (Smithers and Caldecott, 2004). The equation written in finite difference form is expressed as: Sn+1 – Sn = (ln + ln+1)∆t/2 – (Qn + Qn+1) ∆t/2


2

90


Where: Sn In Qn t

= = = =

channel or temporary storage (m3) at time increment = n inflow rate (m3.s-1) at time increment = n outflow rate (m3.s-1) at time increment = n routing period (s).

The subscripts (n) and (n+1) refer to the number of increments in time interval t. To route a hydrograph through a non-linear reservoir, the storage, outflow relationship and the continuity equation (Eqn 2) are combined to determine the outflow and storage at the end of every time step. (2Sn+1)/∆t + Qn+1 = ln + ln+1 + (2Sn/∆t – Qn)

3

Other than in Southern Africa (South Africa, Botswana, Namibia, Lesotho, Swaziland and Zimbabwe), the model had been applied internationally in research in Botswana Chile, Germany Lesotho, Namibia Swaziland and the US (Shulze et al, 2004). 2.2

HEC-HMS Model

HEC-HMS (Scharffenberg and Fleming, 2010) which is the acronym for Hydrologic Engineering Centre’s Hydrologic Modelling System (HEC-HMS) is hydrologic modelling software developed by the US Army Corps of Engineers Hydrologic Engineering Centre (HEC). It is designed to simulate the precipitation runoff processes of dendritic watershed systems in a wide range of geographic areas such as large river basins and small urban or natural watersheds (Scharffenberg and Fleming, 2010). The system encompasses losses, runoff transform, open channel routing, and analysis of meteorological data, rainfallrunoff simulation, and parameter estimation. HEC-HMS uses separate models to represent each component of the runoff process, including models that compute runoff volume, models of direct runoff, and models of base flow. Each model run combines a basin model, meteorological model, and control specifications with run options to obtain results. The system connectivity and physical data describing the watershed are stored in the basin model. The precipitation data necessary to simulate watershed processes are stored in the meteorological model (Kumar et al, 2011). HEC-HMS includes models of infiltration from the land surface but it does not model storage and movement of water vertically within the soil layer. It implicitly combines the near surface flow and overland flow and models this as direct runoff. HEC-HMS considers that all land and water in a watershed can be categorized as either directly connected impervious surface or pervious surface. The curve number method provides relationships between initial abstractions, Ia, and curve numbers, CN, based on experiments carried out in small experimental watersheds. The equations are presented as: S = 1000/CN – 10

4

Ia = 0.2S

5

Also, a relationship for excess rainfall has been established as: Pe = (P – Ia)2/(P – Ia + S)

6

Where S is potential maximum retention in inches, P is the total precipitation in inches and e P is excess precipitation in inches. The curve number is varying from 0 to 100. The curve number is zero for perfectly pervious surfaces and thus Q = 0. The curve number is 100 for perfectly impervious surfaces and thus Q = P. HEC-HMS transforms the rainfall excess into direct surface runoff through a unit hydrograph or by the kinematics wave transformation. In the present study, SCS unit hydrograph (SCS UH) model has been applied for estimating direct runoff. Research by the SCS suggests that the UH peak (UP) and time of UH peak (TP) are related as:


91


qp = CA/Tp

7

Where C= 483.4 in English system, and A is the drainage area square miles. Tp is expressed as: Tp = tr/2 + tlag

8

Where, tr is the excess rainfall duration in hours and tlag is the basin lag time in hours. The basin lag time is defined as the difference in time between the centre of mass of rainfall excess and the peak discharge of the unit hydrograph. The time parameters used in the models were time of concentration and sub basin lag time. 9 Where Tlag is equal to the lag time (in hours) between the centre of mass of rainfall excess and the peak of the unit hydrograph, L is the watershed length in m, CN is the curve number (dimensionless) and Y is the watershed slope in percent (HEC, 1998). 2.3

Catchment Description

Agbogbo catchment (Fig 2) of Ile-Ife, a city in Southwest Nigeria is at the intersection of Latitude 7032’N and Longitude 4032’E (Ogunkoya, 2000). Agbogbo stream has a basin area of 0.4 km2 and a perimeter of 3630.8m that is underlain by the Pre-Cambrian Basement Complex bounded by elongated inselbergs. Soil in the drainage basin reflect the underlying geology and is shallower than 2m. The climate in the drainage basin consists of two seasons: the dry season, extending from November to March, and the wet season, from April to October. Temperatures in the dry season range from a night-time mean of 210C to a daytime mean of 300C and the catchment is covered mainly by farms planted to a variety of tropical food and tree crops (Ogunkoya, 2000).

(Scale 1:20000) Fig. 2: Digital Elevation Model of the Catchment


92


2. 4 Data Collation Method Rainfall data for Agbogbo catchment was collected using the Dines tilting syphon rain recorder while current meter was used to obtain streamflow discharge data. The digital elevation model of the catchment developed from the contour map of Agbogbo catchment was also obtained and streamflow discharge records from January 1987 to March 1990 were used for model calibration and validation. Streamflow data for the 1st hydrologic year (April 1987 – March 1988) were used to calibrate the models by adjusting parameters in the ACRU and HEC-HMS models independently to achieve reasonable agreement between the predicted and observed streamflow. By reasonable agreement is meant an order of magnitude correspondence between the simulated and the recorded series, which is consistent within the duration of an event (Mbajiorgu, 1995). After several simulations, the values of these parameters that provide the closest match and similarity in the hydrographs of the simulated and the observed streamflows for the 1st hydrologic year were used in simulating flows for the 2nd and 3rd hydrologic years (i.e. April 1988 – March 1989 and April 1989 – March 1990) For the ACRU Model, Agbogbo catchment was divided into 3 subcatchments (Agb1, Agb2 and Agb3) with areas 0.1, 0.1, and 0.2km2. In applying the HEC-HMS model, the catchment was divided into ‘Basin 1’, ‘Basin 2’ and ‘Basin 3’with areas 0.1, 0.1, and 0.2km2. Streamflow is routed from the subcatchment Agb1 (Basin 1 in the case of HEC-HMS) down through Agb2 (Basin 2 for HEC-HMS) to the stream outlet (at which point discharge measurements were taken) at Agb3 (i.e. Basin 3) as depicted in Fig 3.

Fig. 3: Sub-catchments Configuration of Agbogbo Catchment 3.


After the calibration run (April 1987 – March 1988), ACRU Model over-predicted streamflow (Fig. 4) for the two successive hydrologic years (April 1988–March 1989 and April 1989–March 1990). The timing of the peaks for the observed and simulated streamflow during the simulation run however coincide which demonstrates that the model is sensitive to changes within the catchment. The Nash–Sutcliffe model efficiency coefficient (Nash et al., 1970) which is a measure of the predictive power of hydrological models and defined as;

∑ T (Qto Qtm)2 E=1

t=0 ∑ T (Qto t=0

Qo)2

was also used to compare the observed discharge (Qo) and the predicted discharge (Qm) at time t. The model efficiency coefficient E, for the ACRU model simulation was obtained to be 0.83695 for the 2 nd hydrologic year (April 1988 to March 1989) while simulation by HEC-HMS model for the same period yielded 0.87748. For the 3rd hydrologic year (April 1989 to March 1990), ACRU model simulation Nigerian Institution of Agricultural Engineers © www.niae.net

93


yielded a model efficiency coefficient of 0.78232 while that of HEC-HMS yielded 0.54706 (full simulation result is given in Appendix 1). Overall (April 1988 to March 1990), ACRU model performed better with a model efficiency coefficient of 0.81218 when compared with the simulation performed by HEC-HMS which gave a value of 0.71498. The hydrograph of the observed and simulated streamflow (Fig. 4) also shows that HEC-HMS is less sensitive to changes within the catchment when compared with ACRU model is illustrated by the mismatch in the timing of the peaks of the hydrograph.

Fig. 4: Comparison between Simulated and Observed Hydrographs The similarity in the occurrence of the timing in the peaks of the hydrographs of the ACRU model simulated streamflow and the observed streamflow shows that the model can be trusted to give reliable flood forecast for the catchment. 4.

CONCLUSIONS

Data obtained from the field and those generated from a digital elevation model of the catchment were used in modelling streamflow for Agbogbo catchment. Data obtained for the 1 st hydrologic year April 1987–March 1988 was used to calibrate the ACRU model and HEC-HMS while data for the next two hydrologic years (April 1988 to March 1990) was used for model validations. When Nash–Sutcliffe Model Efficiency Coefficient was used for the analysis. ACRU Agrohydrologic Model performed better than HEC-HMS over the two years and also showed significant sensitivity to changes within the catchment by the similarity in the timing of the occurrence of peaks for the observed and simulated hydrographs. However, the 3-year duration of the data used for this study (which were the only complete set of data available at the time this study was conducted) is a limitation to the validation exercise. REFERENCES Beven, K. 2001. Rainfall-runoff Modeling – The Primer. John Wiley and Sons, Ltd, Chichester, UK, pp. 1–69. Downing T.E., Ringius L, Hulme M, Waughray D., 1997. Adapting to Climate Change in Africa. Mitigation and Adaptation Strategies for Global Change 2: 19-44, 1997 Kluwer Academic Publishers Belgium. Nigerian Institution of Agricultural Engineers © www.niae.net

94


Gleick, P.H., 1996. Water resources. In Encyclopedia of Climate and Weather, ed. by S. H. Schneider, Oxford University Press, New York, vol. 2, pp.817-823. Gordon, W.S., Famiglietti, J.S., Fowler, N.L., Kittel, T.G., Hibbard, K.A., 2004. Validation of Simulated Runoff from Six Terrestrial Ecosystem Models: Results from Vemap. Ecological Application, 14(2), 2004, p 528. Ecological Society of America. Hobson, A.N., 1997. Use of a Stochastic Weather Generator in a Watershed. An M.Sc thesis submitted in the Department of Civil, Environmental and Architectural Engineering, University of Colorado Hugh, D.A., 2002. Modelling Semi-Arid and Arid Hydrology and Water Resources – The Southern African Experience. IEMSS (2002) Proceedings (http://www.iemss.org/iemss2002/proceedings/ pdf) Kumar D, Bhattachajya R.K., 2011. Distributed Rainfall Runoff Modelling. International Journal of Earth Sciences and Engineering, Volume 04, No. 06 SPL, October 2011, pp. 270-275 Mbajiorgu, C.C., 1995. Watershed Resources Management (WRM) Model 2. An Application to the Upper Wilmot Watershed. Computers and Electronics in Agriculture Vol.13, No.3, pp 217–226, ELSEVIER Muthukrishnan S., Harbor J., Lim K.J., Engel B.A. 2006. Calibration of a Simple Rainfall-Runoff Model for Long-term Hydrological Impact Evaluation. URISA Journal Vol. 18, No. 2 New, M., 2002. Climate Change and Water Resources in the Southwestern Cape, South Africa. South African Journal of Science 98, Month 2002 Ogunkoya, O.O., 2000. Discrepancies in Discharge Records Derived Using the Staff Guage, Staff Guage– Crest Stage Indicator and Water Level Recorder in SW Nigeria. The Nigerian Geographical Journal (New Series) Vol 3 and 4 (2000) pp 169–182 Scharffenberg W.A. and Fleming M.J. 2010, Hydrologic Modeling System HEC-HMS, User’s Manual http://www.hec.usace.army.mil/software/hec-geohms/ Schmidt, E.J., Schulze, R.E., 1987. Flood Volume and Peak Discharge from Small Catchments in Southern Africa, based on the SCS Technique. Water Research Commission, Pretoria, Technology Transfer Report TT/3/87. Schulze, R.E. (ed.) 1989. ACRU: Background Concepts and Theory. ACRU Report No. 36, Dept. Agric. Eng., Univ. of Natal, Pietermaritzburg, RSA. Schulze, R.E. 1994. Hydrology and Agrohydrology: A text to accompany the ACRU 3.00 Agrohydrological Modelling System. Water Research Commission, Pretoria, Report. Schulze, R.E., Angus G.R., Lynch S.D. and Smithers J.C., 2004. ACRU: Concepts and Structure. ACRU Agrohydrological Modelling System (User Manual 4.00 Chapter 2). Smithers, J.C., Dent M.C., Lynch, S.D., Schulze, R.E., 1994. Preparation of Daily Climate Input Files. Chapter four, ACRU 3.00 User Manual. United Nations 2009. Climate Change and Water. An overview from the World Water Development Report 3: Water in a Changing World.. Published by the United Nations World Water Assessment Programme, Programme Office for Global Water Assessment, Division of Water Sciences, UNESCO. 06134 Colombella, Perugia. Whitfield, P.H., Wang, J.Y., Cannon, A.J. 2003. Modelling Future Streamflow Extremes – Floods and Low Flows in Georgia Basin, British Columbia Canadian Water Resources Journal Vol. 28 (4): 633 – 656.


95


APPENDIX Comparison of the observed and the simulated flow for the period under study Date Observed Streamflow ACRU Simulated HEC-HMS Simulated Streamflow Streamflow Apr-87 1.54 2.7 1.8 May-87 1.96 5.3 2 Jun-87 5.98 7.3 6 Jul-87 15.43 21.3 16.4 Aug-87 53.69 61.2 55 Sep-87 106 107.7 111 Oct-87 117.81 117.43 115.6 Nov-87 39.2 47.55 40 Dec-87 21.11 19.33 20 Jan-88 9.06 7.47 8.8 Feb-88 4.88 2.7 5.2 Mar-88 3.54 4.7 4.1 Apr-88 6.38 6.3 5.3 May-88 15.95 15.7 11.7 Jun-88 60.65 47 57.8 Jul-88 68.8 84.6 83.2 Aug-88 76.2 94 99.4 Sep-88 98.7 122.1 123 Oct-88 137.9 140.3 139.5 Nov-88 112.2 140 144.3 Dec-88 35.1 74.9 55.3 Jan-89 12.78 26.2 22.3 Feb-89 6.33 9.4 7.1 Mar-89 4.78 5.8 5.2 Apr-89 6.33 5.6 3.3 May-89 9.58 5.8 7 Jun-89 15.95 11.6 10.3 Jul-89 79.78 59.6 65.4 Aug-89 122 136.8 88.4 Sep-89 113.6 135.4 122.3 Oct-89 94.5 126.8 143 Nov-89 41.5 95.9 122 Dec-89 8.8 16.9 32 Jan-90 4.78 7.1 13.5 Feb-90 1.6 5.1 1.6 Mar-90 6.7 2.2 6.5


96


INFILTRATION THROUGH TRAFFIC COMPACTED SOIL I. E. Ahaneku Department of Agricultural and Bioresources Engineering, Michael Okpara University of Agriculture, Umudike, P.M.B.7267 Umuahia, Abia State, Nigeria. E-mail: [email protected] ABSTRACT Infiltration tests were conducted on a sandy loam soil subjected to three levels of compaction using 5, 7, and 12 tractor passes with a tractor weighing 40 kN. Results indicated that 5 tractor passes induced average soil bulk density of 1.72 Mgm-3, hydraulic conductivity of 1.74 x 10-6 cm/s, and penetration resistance of 479.55 kPa at a moisture content of 14.91 %. The 7 tractor passes produced average bulk density of 1.81 Mgm-3, hydraulic conductivity of 1.24 x 10-5 cm/s and penetration resistance of 514.7 kPa at a moisture content of 12.54 %, while the 12 tractor passes produced average bulk density of 1.91 Mgm3 , hydraulic conductivity of 1.69 x 10-5 cm/s, and penetration resistance of 474.35 kPa, at a moisture content of 12.11 %. The findings revealed that the number of tractor passes imposed did not significantly (p> 0.05) influence water infiltration into the soil. However, soil strength increased with compaction without a corresponding decrease in infiltration. It is concluded that soil type, soil moisture content and machinery capacity are crucial factors in the infiltration process. KEYWORDS: Tractor capacity, tractor passes, soil strength, soil water intake. 1.

INTRODUCTION

The quest and urge to make agriculture less rigorous, more interesting and to reduce the drudgery associated with agricultural practices gave rise to the mechanization of agriculture. As useful and timeefficient as agricultural machinery and equipment may be in the execution of field operations, the heavy weight of these equipment results to soil compaction which has adverse effects on the rate of water infiltration into soils. Compaction increases the bulk density of soil due to a decrease in the number and volume of large pores, which in turn alters aeration, water infiltration, and hydraulic conductivity, and increases soil strength (Blouin et al. 2008). The detrimental effects of compaction are not universal, and are affected by soil type (Gomez et al. 2002) and the level of compaction (Sanchez et al. 2006). Low soil oxygen level caused by soil compaction is the primary factor limiting plant growth. Soil compaction is difficult to correct, being the resulting effect from field traffic of farm machinery. Excessively compacted soil results in problems such as poor root penetration, reduced internal soil drainage, reduced rainfall infiltration, and lack of soil aeration from larger macro pores. According to Soanne and Van Ouwerkerk (1994), soil compaction is responsible for the degradation of an estimated 83 million hectares of land worldwide. Traffic over the soil is the major contributor to soil compaction e.g. a moist soil could reach 75% maximum compaction the first time it is stepped on and 90% by the fourth time it is stepped on (Whiting et al. 2010). According to FAO (2008), soil compaction can lead to reduction in the pore size and spaces through which air and water flow through the soil thereby leading to runoff, erosion and causing emergence of agricultural wastelands. Studies have shown that most soil compaction at the urban lot scale leads to increased storm water runoff and this is particularly crucial in low impact development strategies where storm water is intended to infiltrate rather than flow through a traditional network to a detention basin (Pitt et al. 2002). Therefore, identifying the presence of compaction is crucial to understanding its effect on the infiltration rate and relating such to crop yield and environmental problems of runoff and erosion.


97


Soil can be an excellent temporary storage medium for water, depending on the type and condition of the soil. Infiltration rate and cumulative infiltration are two parameters commonly used in evaluating the infiltration characteristics of a soil; infiltration being a dynamic process, variable in time and space plays a vital role in the replenishment of soil water which is responsible for the growth and development of crops (Ahaneku, 2011). Decrease in infiltration or increase in saturation above a compacted layer can also cause nutrient deficiencies in crops. Either condition can result in anaerobic conditions which reduce biological activity and fertilizer use efficiencies (Duike 2004). There has been relatively little research done on the effects of traffic-induced compaction on water infiltration in Nigerian soil. The essence of this study was to determine the effects of different levels of compaction by tractor passes on infiltration rates and physical properties of a sandy loam soil. 2.


2.1

Description of Study Site

The study site is the Federal University of Technology, Gidan Kwano campus, Minna, Niger state, Nigeria. The site was relatively undisturbed as no cultural or tillage activities had been carried-out on it for six months prior to the experiment. Minna is in the southern Guinea Savannah vegetation of Nigeria (latitude 09 0 31.8’ N and longitude 06 0 27.1’E). Niger state lies in the semi-arid zone and has two distinct seasons, wet and dry season. The wet season starts in April and ends in October with the maximum annual rainfall recorded in August. Minna has a mean annual rainfall of 1220 mm. The average maximum and minimum temperature for Minna is 31 0C and 28 0C, respectively with annual relative humidity of 59 %. 2.2

Specifications of Tractor Used

The specifications of the tractor used to achieve compaction is given in Table 1 Table 1.Tractor specifications Tractor Model Engine Type Weight Effective output Tyre inflation pressure Speed of operation Gear selection 2.3

Steyr 8075 4-cylinder diesel engine Maximum gross weight: 4000kg Maximum gross axle weight: 1500kg 64hp (47.7kW) Front wheel: 275.79 kPa Rear wheel: 103.42 kPa 10km/hr Gear 2 (high)

Compaction Test

The experimental site was divided into three plots for the purpose of the compaction test. The treatment consisted of three levels of compaction: 5 tractor passes, 7 tractor passes, and 12 tractor passes, with one level imposed on each of the three plots. A tractor of known specification was used to achieve the compaction by driving it between two points at the specified number of passes.


98


2.4

Determination of Infiltration rate

Prior to the tests, soil samples were collected from the site and taken to the laboratory for determination of soil physical properties like bulk density, particle density, moisture content, porosity and the soil textural class. Infiltration tests were carried-out on each level of compaction at varied time intervals within a two-hour frame with three (3) replicates at each level of compaction. The infiltration measurement was done using double ring infiltrometers. The inner ring of the infiltrometer is 34cm in diameter, while the outer ring has a diameter of 65cm with a height of 25cm. The rings were inserted into the soil and left overnight and the infiltration tests carried out the next day. The field data was used to determine the infiltration rate, cumulative infiltration and the average infiltration rate of the soil. 2.5

Determination of Soil Penetration Resistance

Soil penetration resistance was measured in each plot in three replicates using a hand-held penetrometer. The shank with a cross-sectional area of 4.84 cm2 was used for 5 passes level of compaction, while shank with a cross-sectional area of 6.45 cm2 was used for 7 passes and 12 passes compaction, respectively. 2.6

Soil Texture and Moisture Content determination

Soil texture was determined by the hydrometer method, while soil moisture content was determined gravimetrically. Soil samples were collected from the field after compaction and taken to the laboratory; the soil was weighed and placed in the oven and heated at 105 0C for 24 hours after which it was weighed again. The change in mass represents the moisture content. 2.7

Soil Bulk Density Determination

Soil bulk density was determined from oven dried undisturbed cores as mass per volume of oven dried soil. 2.8

Soil Porosity Determination

Porosity was determined using the relationship between bulk density and particle density as: 𝐵𝑢𝑙𝑘 𝑑𝑒𝑛𝑠𝑖𝑡𝑦

% Porosity = 1 − [𝑃𝑎𝑟𝑡𝑖𝑐𝑙𝑒 𝑑𝑒𝑛𝑠𝑖𝑡𝑦] x 100 2.9

(1)

Determination of Field Capacity

Three 65 cm infiltrometer rings were used to pond water on each of the compacted sites and left undisturbed for 24 hours. This was done to allow the water within the ring to drain through the soil profile completely after which soil samples were taken from each of the locations and oven-dried for 24 hours to determine the average moisture content. 2.10 Determination of Hydraulic Conductivity The hydraulic conductivities of the soil at the three levels of compaction were determined in the laboratory using constant-head permeameter method (Israelsen and Hansen, 1962).


99


2.11

Data Analyses

The SPSS IBM version 20 software was used for the statistical analyses. To test the equality of means of the treatment parameters, one-way analyses of variance (ANOVA) was applied and mean comparison evaluated using Chi-square test. 3.


3.1

Effect of level of Compaction on Infiltration and Soil Physical Properties

The soil particle size distribution revealed the soil texture at the experimental site to be sandy loam (Table 2). Infiltration tests carried out on 5 tractor passes plot showed that final infiltration rate ranged from 1.5 mm/hr to 2.75 mm/hr. Soil penetration resistance of between 466.20 kPa and 492.50 kPa was observed with average soil moisture content of 14.91 % and bulk density of 1.72 Mgm-3. Observed infiltration rate for the 7 tractor passes ranged from 7.75 mm/hr to 17.5 mm/hr. The penetration resistance was found to be between 509.5 kPa and 520.2 kPa at an average soil moisture content of 12.54 % and bulk density range of 1.81 Mgm-3. Soil compaction for the 12 tractor passes yielded final infiltration rate ranging from 12 mm/hr to 18.75 mm/hr. The soil penetration resistance was found to be between 471.5 kPa and 477.2 kPa at average moisture of 12.11 % and bulk density of 1.91 Mgm-3. Table 2.Particle size distribution of test soil Site Compaction Level Sand (%) A 5 Passes 68.62 B 7 Passes 71.52 C 12 Passes 66.12

Silt (%) 17.90 16.40 20.20

Clay (%) 13.48 12.08 13.68

Textural Class Sandy Loam Sandy Loam Sandy Loam

3.2 Impact of Compaction on Infiltration and Soil Properties The values of the control parameters are presented in Table 3. Comparison of means of soil properties are shown in Table 4, while the average infiltration rates and cumulative infiltration are shown in Figs.1 and 2, respectively. As can be seen from Table 4, no significant difference was found in all the parameters with the treatments. The result of the infiltration tests showed a lot of variability, with the lowest level of compaction (5 tractor passes compacted soil), having the least infiltration rate of between1.5 to 2.75 mm/hr, while the highest level of compaction (12 tractor passes) exhibiting higher infiltration rate of between 12 mm/hr to 18.75 mm/hr. The observed cumulative infiltration for the 5 tractor passes compacted site ranged between 3 mm to 5.5 mm, while those for the 7 and 12 tractor passes were 15.5 mm to 35 mm and 24 mm to 37.5 mm, respectively. The median initial infiltration rate for 5 passes compacted soil is 3.5 mm/hr. The 7 tractor passes compacted soil had initial infiltration rate ranging from 10.5 mm/hr to 22 mm/hr, while that of 12 passes compacted soil ranged from 16 to 22 mm/hr. This shows that within the first hour, the 12 passes compacted soil had a higher initial infiltration rate than the 5 and 7 passes compacted plots which was not anticipated. Table 3. Values of control parameters MC FC IR HC (%) (%) (mmhr-1) (×10-6 cms-1) 18.24 22.44 21.47 1410.00 MC = Moisture content, FC= Field capacity, IR Penetration resistance, BD = Bulk density

PR BD Porosity (%) (kPa) (Mgm-3) 349.40 1.34 41.50 = Infiltration rate, HC = Hydraulic conductivity, PR =


100


Table 4. Comparison of mean soil properties and their standard deviations in the studied treatments Number of MC IR HC x 10-6 PR BD Porosity -1 -1 -3 Tractor (%) (mm hr ) (cm s ) (kPa) (Mg m ) (%) Passes 5 14.91±2.23 2.17±0.14 1.74±0.03 479.55±45.01 1.72±0.04 34.34±0.30 7 12.54±2.00 12.00±0.19 12.40±0.51 514.70±13.90 1.81±0.03 31.70±0.01 12 12.11±1.08 15.92±0.74 16.90±0.11 474.35±2.28 1.91±0.04 27.92±0.05 Sig. 0.076 0.083 0.072 0.060 0.063 0.56 (2-tailed) MC = Moisture content, IR = Infiltration rate, HC = Hydraulic conductivity, PR = Penetration resistance, BD = Bulk density

Average Infiltration Rate (mm/hr)

60 50 40

5 Passes

30

7 Passes

20

12 Passes

10 0 5

10 15 30 45 60 75 90 105 120 Time (mins)

Average Cummulative Inflltration (mm)

Fig. 1. Average infiltration rates for 5, 7, 12 Passes

35 30 25 20

5 passes

15

7 passes

10

12 passes

5 0 0

20

40

60

80

100

120

140

Time (mins)

Fig. 2. Average cumulative infiltration for 5, 7, and 12 Passes The observed results do not mean that compaction is a favourable phenomenon which enhances infiltration. Table 2 shows that there was more silt in the plot with 12 tractor passes than in the other two plots. The higher infiltration rate observed in the 12 tractor passes could probably be due to the heterogeneity of the soil (having more sand and silt in the lower horizon of the 12 tractor passes plot than in the other two plots). Also, compaction at near field capacity as was the case with the 5 tractor passes could reduce infiltration rates (Akram and Kemper, 1979). Nigerian Institution of Agricultural Engineers © www.niae.net

101


According to Pitt et al. (2001), a more detailed testing could explain some of the large variations observed. During the first hour it was also observed that there was no movement of water at certain intervals which may be due to subsurface compaction that occurred during the tractor passes or as a result of activities performed on the land years before the tests, as a reliable history of the site, two1 or more years prior to the tests was not available. After two hours of testing, the average final infiltration rates were 2.25, 12 and 15.92 mm/hr for the 5, 7 and 12 tractor passes compacted plots, respectively. The area with the poorest drainage (5 tractor passes compacted plot) had the least infiltration rate. The 12 tractor passes compacted plot showed the highest infiltration rate for the three replicates (17, 18.75 and 12 mm/hr). Water infiltration into the soil was more during the first hour than in the second hour for all the replicates which conformed to literature (Siyal et al. 2002) that more water will be infiltrated in the early minutes of an infiltration test and as the time increases, the amount of water infiltrated reduces until a constant water level is attained. The compaction at the three locations was found to reduce the infiltration rate from 21.47 mm/hr before compaction to an average of 2.17 mm/hr for the 5 passes making the soil behave like clay with a steady infiltration rate ranging from 1 to 5 mm/hr as shown in Fig.1. Average infiltration rates of 12 mm/hr and 15.92 mm/hr in the 7 tractor passes and the 12 passes, respectively showed infiltration rates characteristics of loam (Fig.1). The variability and inconsistency observed during the infiltration tests can be reduced if more tests are conducted and a smaller diameter infiltrometer ring used as recommended by Pitt et al. (2001). Penetration resistance of soils is dependent on the moisture content of the soil. The values for the compacted soils were very high and may impact negatively on root penetration. The bulk densities for the three compacted sites at 5, 7, and 12 tractor passes were found to have average values of 1.74 Mgm-3, 1.81 Mgm-3, and 1.91 Mgm-3, respectively, in relation to 1.34 Mgm-3 for the control location. Infiltration rate was affected by the high bulk densities observed. In spite of this, the 12 tractor passes compacted soil showed the highest infiltration rate with an average value of 15.92 mm/hr. The observed results do not suggest that the more the traffic on soils, the less the infiltration rate. Similar findings were reported by Whiting et al. (2010). The observed results may be due to the fact that the soil was compacted within accepted moisture range for sandy loam soil (less than 28.2% dry basis) and the tractor capacity used was within the established safe loads (60kN) for sandy loam fields as reported by Kanali et al. (1997). Notwithstanding the observed infiltration figures, the high bulk densities may limit root penetration in accord with The Cooperative Soil Survey (2011) report that bulk density from 1.6 Mgm-3 will have severe negative impact on plant growth. 3.3

Compaction, Soil Moisture and Hydraulic Conductivity Relationship

As with infiltration rate, the hydraulic conductivities obtained at the three levels of compaction also indicate that the 5 tractor passes compacted soil with bulk density of 1.74 Mgm-3 had the least value of hydraulic conductivity ( 1.74 x 10-6 cm/s ), while 7 tractor passes with bulk density of 1.81 Mgm-3 had a value of 1.24 x 10-5 cm/s, and the 12 tractor passes with bulk density of 1.91 Mgm-3 had a value of 1.69 x 10-5 cm/s. According to Kooistra et al. (1984), the higher the bulk densities of soils, the lesser the hydraulic conductivity, but this is not always true because clay with a bulk density of about 1.4 Mgm-3 can have hydraulic conductivity of between 10-9 to 10-6 cm/s (The Cooperative Soil Survey, 2011). The 5 tractor passes compacted soil with soil moisture of 14.91 % had the highest value of conductivity closest to the field of capacity of the soil (23.44 %), indicating that compaction was carried-out near field capacity leading to the reduced infiltration rate observed in that treatment with an average rate of 2.17 mm/hr after 120 minutes. The 7 tractor passes compacted soil with moisture content at 12.54 % had a higher observed average infiltration rate of 12 mm/hr after two hours. The 12 tractor passes with the least Nigerian Institution of Agricultural Engineers © www.niae.net

102


moisture content (12.11 %), had the highest average infiltration rate of 15.92 mm/hr at the end of two hours. The results show that soil moisture content before compaction strongly affected soil strength properties. The outcome of this study is in agreement with the findings of Blouin et al. (2008) in a similar study. 4.

CONCLUSIONS

The study evaluated the effects of traffic compaction on water infiltration and soil physical properties of a sandy loam soil. Results of the study showed that in all parameters measured, no significant difference was found between the treatments. However, there was appreciable increase in soil strength properties (bulk density and penetration resistance) with compaction. The findings of this study suggest that provided the soil is not too wet and the axle load of the machinery used in the field is within established limits for a particular soil type, compaction may not limit infiltration adversely within the limits of this experiment). With the results of this study, farm managers may make informed decisions on the effects of traffic compaction on soil productivity. ACKNOWLEDGEMENT: The author would like to acknowledge the assistance of Mr. Odewale, Babajide Qosim during the field work. REFERENCES Ahaneku I.E., 2011. Infiltration characteristics of two major agricultural soils in north central Nigeria. Agricultural Science Research Journals 1(7): 166-171 Akram, M. and Kemper W. D. 1979. Infiltration of soils as affected by the pressure and water content at the time of compaction. Soil Service Society American Journal. 43: 1080-1086. Blouin V.M., Schmidt M.G., Bulmer C.E., Krzic M., 2008. Effects of Compaction and Water Content on Lodgepole Pine Seedling Growth. Forest Ecology and Management 255:2444-2452. Duiker S. W., 2004. Effects of soil compaction. Retrieved from www.cas.psu.edu (accessed June 6, 2011). FAO, 2008. Retrieved from http://www.fao.com/landandwater. (accessed July 7, 2011). Gomez A., Powers, R.F., Singer, M.J., Howarth, W.R., 2002. Soil Compaction Effects on Growth of Young Ponderosa Pine Following Litter Removal in California’s Sierra Nevada. Soil Sci. Soc. Am. J. 66, 1334-1343. Israelsen O.W.and Hansen, V.E., 1962. Irrigation Principles and Practices. John Wiley and Sons, Inc., New York. Kanali C. L, Kaumbutho P.G., Maende C.M. and Kamau,J. 1997. Journal of Terramechanics, Vol. 34, No. 2, pp. 127-140. Kooistra M.J., Bouma J., Boersma O.H., Jager A., 1984. Physical and morphological characterisation of undisturbed plough pans in a sandy loam soil. Soil Tillage Research 4:405-417 Pitt R., Shen-En Chen, Shirley C., 2001. Infiltration through compacted urban soils and effects on biofiltration design. Presented at the Low Impact Development Roundtable Conference, Baltimore, MD, July, 2001. Pitt R., Shen-En Chen, Shirley C., 2002. Compacted urban soils effects on infiltration and bioretention stormwater control designs. Presented at the 9th International Conference on urban Drainage. IAHR, IWA, EWRI and ASCE. Portland Oregon, September 8-13, 2002 Sanchez F.G., Scott, D.A., Ludovici, K.H., 2006. Negligible Effects of Severe Organic Matter Removal and Soil Compaction on Loblolly Pine Growth over 10 Years. For. Ecol. Manage. 227, 145-154. Siyal A.G., Oad, F.C., Samo M. A., Zia-Ul-Hassan, Oad N.L., 2002. Effect of compactions on infiltration characteristics of soil. Asian Journal of Plant Sciences. 1 (1): 3-4. Soanne B. D., Ourwerkerk C. Van., 1994. Soil compaction problems in world agriculture. Scottish Centre of Agricultural Engineering, SAC, Penicuik, U.K. and Institute for soil fertility research (IB-DLO), Haren Gn, Netherlands. pp 2


103


The Cooperative Soil Survey, 2011. Soil Bulk Density-Physical Properties. Retrieved from soilsurvey.org accessed on 13 November, 2011. Whiting D, Adrian C, Carl W Reeder J., 2010. CMG Garden Notes #215. Colorado State University Extension. Retrieved from www.cmg.colostate.edu (accessed June 6, 2011).


104


INVESTIGATION OF GROUND WATER QUALITY DURING THE DRY SEASON IN RIVERS STATE, NIGERIA: A CASE STUDY OF PORT HARCOURT METROPOLIS I. Fubara-Manuel and R. B. Jumbo Department of Agricultural and Environmental Engineering, RSUST, Port Harcourt, Nigeria. [email protected] ABSTRACT This research is to investigate the quality of ground water in boreholes and hand-dug wells during the dry season in six locations in Port Harcourt metropolis of Rivers State. The locations are Rumuola, D-line and Trans-Amadi, which use borehole while Elechi Beach, Gbundu and Ozubuko still use hand-dug wells. The parameters analysed were iron, chloride, hardness, and pH. Others were total dissolved solids (TDS), ammonia (free and saline), and E-coli. Results indicated that for iron, all except Ozubuko with a value of 0.32mg/l satisfied WHO’s standard. It is not sufficient, however, to reject water containing higher iron content unless a more suitable supply is available. For parameters such as chloride and hardness, the samples from all the locations are within WHO’s limit. This is also true for TDS, while for pH, the water samples from all the locations were acidic, falling short of the standard (6.5 – 8.5) specified by WHO although samples from Rumuola with the highest value of 6.66 fall within the WHO range and hence acceptable. Ammonia levels in the water samples from all the locations were within the recommended value by WHO, with the highest value (5.01) coming from Trans-Amadi, which indicates the highest level of biological reductions in this water. Finally and of greater significance too, is the fact that no E-coli was found in all the samples except that from Gbundu waterside with a level of 3 x 10 3 as against zero recommended by WHO. This therefore makes the hand-dug well water from Gbundu warterside unsuitable for human consumption. KEY WORDS: Ground water, quality, wells, boreholes. 1.

INTRODUCTION

Water is an indispensable resource outside which human existence will become unbearable. Drinking water should be clear, cool, free from objectionable tastes and odours and from harmful chemicals and microorganisms (Noha, 2007). The sources of water in the world include the entire range of natural waters that occur on the earth, which include underground water, surface water and rain water. Surface water include rain water collected from structures or prepared catchments, water from rivers, natural lakes, storage reservoirs and oceans. Underground water supplies are a result of surface water percolating through the soil and rock. They include natural springs, shallow wells, deep and artesian wells and horizontal galleries. Rain water is the purest form of natural water because it is formed as a result of the condensation of water vapour in the atmosphere that is, it is a natural form of distilled water. Spring-water contains a considerable amount of mineral salt, but with very little suspended impurities such as dust and bacteria. Water from deep wells or boreholes tends to be free from pollution while water from shallow wells are prone to contain some forms of contaminants. Improper management of waste especially treatment and disposal of solid and liquid wastes are the major contributors to urban area water pollution (Napacho, 2010). There are various ways groundwater may suffer pollution e.g. land disposal of solid wastes, sewage disposal on land, Agricultural activities, urban runoff and polluted surface water (Jain et al, 1995). The quality of ground water depends on the quality of water recharging the aquifer and the hydrologic and biogeochemical processes that affect it along flow paths from recharge to discharge areas (Kevin et al., 2004). Problems with drinking water infrastructure, whether public or private, threaten the safety, quality and health values of drinking water for the public (Eftila, 2010). Most water before they reach the


105


consumer, have been exposed to greater or lesser amount of contamination. Precipitation carries an appreciable amount of materials to the Earth’s surface. Windblown spray from seawater is the sources of many salts that are dissolved in rain and they contain chloride (CL, sodium (Na+), magnesium (Mg2+), calcium (Ca 2+) and potassium (k+) irons. Winds pick up small particles and living organisms such as pollen grains, bacteria and the spores of fungi. These air bone particles are trapped by raindrops and are suspended in them as they fall. The amount and types of impurities in precipitation vary with location and the time of the year, and can affect the characteristics of streams, ponds, lakes, reservoirs, oceans and rivers (Bulbul A., 2013). Other constituents and characteristics that make water less acceptable include total dissolved solid content, colour, turbidity, off taste or odour, phenolic substance carbon chloroform extract, ethyl – benzyl sulfonates, and other wastes not readily degradable. Also products such as detergents, artificial fertilizers, and insecticides become pollutants when they get into water supply (Encylopaedia Britannica, UT – ZW V, 19) Apart from direct factors some other factors indirectly affect quality of water. In the case of the underground water supply sources, excessive withdrawals may have an adverse effect on the chemical quality of the supply, such as increasing the content of iron or manganese or total dissolved solids (Culp and Culp, 1974). Port Harcourt is a coastal city and the capital of Rivers State, Nigeria. The rural to urban migration resulting in an ever increasing population has put a lot of pressure on existing water facilities. This has further incapacitated the ability of government to meet the water demand of the populace. It is therefore common place to see women and children with buckets and gerri cans roaming the street in search of potable water, especially during the dry season. This development has given rise to the booming sachet water, often called “pure water” business, as numerous outlets are springing up all over the city. As a consequence, residents suffer from water related diseases and in extreme cases deaths. Records in the Rivers State Ministry of Health show that between 2009 and 2012, there were 305 recorded cases of cholera outbreak with 23 deaths. Furthermore, in the year 2012 there were 1,074 recorded cases of typhoid and paratyphoid patient with 12 recorded death cases. There were also 2,856 reported cases of amoebiasis and a total of 128 deaths. The report also indicated that 10 percent of infant mortality was as a result of diarrhea or respiratory infections, while other diseases such as dysentery were also on the increase. To bridge the wide gap between demand and supply, residents have resorted to indiscriminate drilling of boreholes and shallow wells. Most often, the shallow wells are located near sources of pollution and the quality is rather poor (Ajayi A.A., 2008). The aim of this research, therefore, was to investigate the quality of the most preferred source of water (ground water) during the dry season in six locations in Port Harcourt metropolis. 2.


2.1

Description of Study Area

Port Harcourt is located in the Niger Delta region, lying along the Bonny river (an eastern distributory of the River Niger), 66km upstream from the Gulf of Guinea. Port Harcourt with latitude4.750 N and longitude 70 E has a tropical monsoon climate, with average temperature averaging between 250 C - 280 C and rainfall measuring an average of just over 210mm. The investigation of water quality was performed in six locations in Port Harcourt metropolis. The locations are: Elechi beach, Gbundu water side, D-line, Trans Amadi, Ozuboko Community and Rumuola. The high population density areas of Elechi beach, Ozuboko and Gbundu waterside mainly use water from shallow wells as potable water, while the low population density and industrialised areas of Trans – Amadi, Rumuola and D-Line mainly use boreholes and pipe-borne water. Figs. 1 to 3 show


106


some of the hand-dug wells in the three high density areas. Samples of water from each location were collected and transported to the laboratory for analysis.

Fig1: Side view of hand-dug well (with protective cover) in Elechi Beach

Fig. 2: Front view of hand-dug well (without protection cover) in Gbundu water side

Fig. 3a: Front view of hand-dug well (without protective cover in Ozubuko


107


Fig. 3b: Plan view of hand-dug well (without protection cover in Ozubuko 2.2

Sample Collection and Test Procedures

The samples were collected in accordance with the procedure recommended in the standard methods for the examination of water and waste water prepared by the American Public Health Association (APHA). The parameters tested were total dissolved solids, pH, iron, chloride, hardness, ammonia test (free and saline) and bacteriological counts. The total dissolve solid was determined with the gravimetric method (APHA 1995) while the pH was determined with the electronic pH metre. Iron was tested with the ASTM D1068-10 standard method, while chloride was determined with the Argentometric titration method (APHA, 1985). Furthermore, hardness and ammonium-nitrogen were determined by the ASTM D1126-12 standard test method and the phenate method (APHA 1985) respectively, while nitrate measurement was by the Brucine method (APHA, 1979). The bacteriological count was performed with the 9215B pour plate using R2A agar. The turbidity was determined with the ASTM D7726-11 standard method while the dissolved oxygen was measured with the dissolved oxygen meter. Electrical conductivity was measured with the electronic conductivity meter while sulfate and fluoride were measured with the ASTM D516-11 and the ASTM D1179-10 standard methods respectively. Also, magnesium was determined with the ASTM D511-09 standard method while aluminum was tested for with the ASTM D857-12 standard method. Cyanide was determined with the ASTM D2036-09 standard method. Furthermore, taste was determined using flavour threshold test (standard method 2160) while colour was determined with the - ASTM D1209 05 standard method. 3.


The results of the various test (Physical, Chemical, and Biology) at Trans- Amadi, Gbundu waterside, Elechi beach and Ozuboko) are shown in Table 1. Table 1: Result of Parameters Tested Parameters Rumuola D(bore Line hole) (bore hole) Iron 0.06 0.08 Chloride 3.34 4.84

TransAmadi (bore hole) 0.27 19.06

Gbundu Elechi (shallow beech well) (shallow well) 0.19 0.16 17.20 7.80


Ozuboko (shallow well)

WHO

0.32 18.92

0-0.30 250 108


Taste Colour Hardness pH TDS TURBIDITY (NTU) Nitrate (mg/l) DISSOLVED OXYGEN (mg/l) ELECTRICAL CONDUCTIVITY (µs/cm) SULFATES AS SO4(mg/l) Fluoride (F-) (mg/l) Magnesium (Mg+2) (mg/l) Aluminum (Al) (mg/l) Cyanide (CN-) ( μg/l) Ammonia (mg/l) E-coli

no clear 4.40 6.66 171.8 0.00

no clear 5.10 6.30 198.2 0.00

no clear 23.51 5.78 223.7 0.00

no clear 30.91 5.50 154.3 0.24

no clear 18.44 5.48 151.9 0.10

no clear 35.50 5.31 138.1 0.00

INOFFENSIVE Clear 1000 6.5-8.5 500 1.00

0.01 6.4

0.04 6.9

0.01 6.7

0.10 7.9

0.06 7.8

0.02 6.8

50 -

180

192

165

210

200

190

250

1.8

1.6

2.0

5.8

5.2

2.8

400

0.2

0.1

0.5

0.8

0.6

0.2

1.5

2.4

1.8

2.1

3.9

4.2

1.2

50

0.02

0.01

0.04

0.01

0.03

0.01

Nil

0.3

0.1

0.2

0.3

0.1

0.1

50

1.21 0.00

1.26 0.00

5.07 0.00

2.88 3 x 103

1.95 0.00

2.01 0.00

10 Nil

For borehole water in respect of iron, all the values are within the range recommended by WHO, while those from well water (except Ozubuko) are also within the WHO range. Although iron can produce taste and colour in water, there is no basis for the rejection of water containing iron in quantities above the WHO standard especially when other more suitable supplies are not available (Fair et al; 1981). The chloride levels in all the water supplies, both from the boreholes and wells are within the specified WHO limit. The results further indicates that there was neither taste nor colour problem in all the water samples. Turbidity and nitrate were also below the recommended value by WHO, thus making water from all the location potable with respect to these parameters. There is no guideline by WHO on dissolved oxygen (Okechukwu et al, 2013) but water sample from Gbundu shallow well recorded the highest value of 7.9mg/l. Conductivity is the measure of the activity of all dissolved ionized solids in water. From the WHO standard, the conductivity values of all the water samples are below the maximum allowable. Sulfate and the fluoride of all the water samples were also below the recommended values by WHO. Also, magnesium and cyanide of all the water samples were below the range recommended by WHO. Though there is no standard value for aluminum in water analysis by WHO, the values obtained from the water sample showed that the values are very minimal and can be accepted as potable. All the values obtained, in terms of hardness, are also far below WHO’S limit. For pH, all the data obtained were below the range specified by WHO except Rumuola, thus indicating acidity of all the water including Rumuola. This may be attributable, in part, to the high volume of industrial activities in these locations especially gas-flaring which results in acid-rain. Ammonia (free and saline) is said to occur in natural water supply as a result of micro-biological reduction and can also indicate pollution by sewage (Nelson and Nelson, 1973). It therefore implies that Nigerian Institution of Agricultural Engineers © www.niae.net

109


the higher the amount of ammonia, the greater the level of microbiological reduction and/or sewage pollution. From the analyzed samples the levels of ammonia range from 1.21mg/L to 5.07mg/L. One can therefore say that the water sample from Trans-Amadi with the highest ammonia level of 5.07mg/l is where more biological reductions took place. Also there was no trace of E-coli in all the samples tested except for Gbundu. This indicates that all the water sources are suitable for human consumption except Gbundu well water. 4.

CONCLUSIONS AND RECOMMENDATIONS

Parameters whose values in all the locations were within the limit set by WHO are chloride, taste, colour, hardness, total dissolved solids, and turbidity. Others include nitrate, electrical conductivity, sulphates, fluorides, magnesium, cyanide, and ammonia. However, for iron, all locations except Ozuboko satisfied WHO standards while for pH, only water from Rumuola fell within the limit set by WHO. Since there was no standard set by WHO for dissolved oxygen and aluminum, it is safe to assume that the results obtained from the water sample pose no threat to health. E-coli level in Gbundu well water is high, thereby rendering it unsuitable for drinking. On the whole, these results indicates that boreholes are better sources of potable water than shallow wells in Port Harcourt metropolis. It is recommended that the indiscriminate drilling of boreholes by individuals and institutions without coordination from an organized body needs to be properly checked. Reliable and efficient contractors should be engaged in the construction of boreholes. Hand-dug wells for water supply should be accompanied with provision for adequate sanitary facilities. The low levels of chloride and fluoride suggest the necessity for some forms of treatment such as chlorination and fluoridation in the boreholes and wells in Port Harcourt metropolis. REFERENCES Ajayi A.A., Sridhar M. K. C., Adekunle Lola V. and Oluwande, P. A. 2008. Quality of Packaged Waters Sold in Ibadan, Nigeria, African Journal of Biomedical Research, Vol. 11: 251 – 258 APHA 1979. Standard Method for the Examination of Water and Waste Water. 18 th Edition. APHAAWWA-WPCF, Washington DC APHA 1985. Standard Method for the Examination of Water and Waste Water. 18th Edition. APHAAWWA-WPCF, Washington DC APHA 1995. Standard Method for the Examination of Water and Waste Water. 18 th Edition. APHAAWWA-WPCF, Washington DC. Arcadio P., Sincero Sr. and Gregoria A.1996. Environmental Engineering: A Design Approach. Asoke Ghosh Prentice-Hall of India. Brown, R.M., Mc Clelland N.I., Deininger R.A. and O’Connor M. F. 1971. The water quality index – crashing the phychological barrier. A Report of Research Study, university of Michigan. Bulbul Ahmed, H. M., Rasel, Md., Shafi Uddin Miah. 2013. Investigate the river water quality parameters: A case study, American Journal of Civil Engineering. 1(3): 84-90, published online September 30. (http://www.sciencepublishinggroup.com/j/ajce), doi:10.11648/j.ajce.20130103.12 Culp, G.L. and Culp, R.L.1974. New concepts in water purification. Litton Educational Publishing INC, London. Eftila Tanellari, Darrell Bosch, Chair, Kevin Boyle, Bradford Mills, Christopher Parmeter, James Pease. 2010. Essay on the Economics of Drinking Water Quality and Infrastructure, April 29, Blacksburg, Virginia. Ekubo A.T. and J.F.N. Abowei. 2011. Aspects of Aquatic Pollution in Nigeria. Research Journal of Environmental and Earth Sciences 3(6): 673-693. Encyclopaedia Britanica, VT – 2W V 19 Fair, G.M. Geyer, M.C. and Okun, D.A. 1981. Element of water supply and waste water removal. 2 nd edition Wiley and sons, INC New York Fair, M.G. 1966. Water and waste water Engineering Vol. 1 Citin book Company Radnor Pennsylvania Nigerian Institution of Agricultural Engineers © www.niae.net

110


Jain CK, Sharma MK, Bhatia KKS, Seth SM (Natl Inst Hydro, Roorkee 247667, UP). Ground water pollution endemic of fluorosis. Polln Res, 19(4) (2000), 505-509 Kevin F. Dennehy, Jason J. Gurdak, Peter B. McMahon, Jennifer S. Stanton, and Sharon L. Qi. 2004. Sources—Quality of Recently Recharged Water in the High Plains Aquifer Napacho Z.A and Manyele, S.V. 2010. Quality assessment of drinking water in Temeke District (part II). Characterization of chemical parameters, African Journal of Environmental Science and Technology Vol. 4(11): 775-789. Nelson, A. and Neslson, Dk 1973. Dictionary of water and water Engineering. Butterworth and Co, (Publishers) ltd, London Noha Donia. 2007. SURVEY OF POTABLE WATER QUALITY PROBLEMS IN EGYPT, Eleventh International Water Technology Conference, IWTC11 Sharm El-Sheikh, Egypt . Okechukwu M.E., Ogwo C.U., Onuegbu C.U., Mbajiorgu G. I., Ezenne G. I. 2013. Water Quality Evaluation of Spring Waters in Nsukka, Nigeria, Nigeria Journal of Technology, vol32, no. 2: 233240. Sangodoyin, A. Y. 1991. Water quality influence and maintenance of rural Boreholes in Nigeria. Environmental studies, Vol. 37: 97 – 107 Sanodoyin, A. Y. 1990. Fundamentals and trends of water services in a Nigerian Urban settlement. Environmental education and information Vol. 9. Number 4: 184 Tebutt, T. H. Y. 1979. Principles of water quality control. 2nd edition oxford, New York. Pergamon press World Health Organization (WHO), International standards for drinking water, (1982)


111


MODELLING CHANNEL ROUGHNESS COEFFICIENT USING DIMENSIONAL ANALYSIS M.Y. Kasali1 and A.O. Ogunlela2 National Centre for Agricultural Mechanization, Ilorin, Kwara State, Nigeria. E- mail: [email protected] 2 Department of Agricultural and Biosystems Engineering University of Ilorin, Ilorin, Kwara State, Nigeria E- mail: [email protected]

1

ABSTRACT A mathematical model for predicting the roughness coefficient of a channel was developed using dimensional analysis based on the Buckingham’s π theorem. Dimensional analysis was used to establish the relationship between the fluid, flow and channel geometric properties. This relationship was subjected to a multiple regression analysis using the field data obtained from the channels, leading to the prediction of the roughness coefficient. The predicted roughness coefficient was validated with Manning’s roughness equation and a high coefficient of correlation of 0.90 and a coefficient of determination, R 2 of 0.814 between the predicted and Manning’s coefficients were obtained. The analysis of variance (ANOVA) showed that there was no significant difference between them at 5% level of significance. KEYWORDS: Roughness coefficient, manning’s equation, dimensional analysis, Buckingham’s π theorem, dimensionless terms. 1.

INTRODUCTION

The roughness coefficient of a channel indicates the degree of the roughness of a channel surface. It defines the extent of the resistance of the surface to flow. The roughness characteristics of an open channel are widely dependent on the geometric, hydraulic and the flow parameters of the channel. This implies that the variation of roughness coefficient occurs due to many contributing factors such as surface, vegetation, channel irregularity, channel alignment, silting and scouring, obstruction as well as discharge of the channel. The determination of the roughness coefficient presents a provocative and a creative task of the contemporary hydraulics of open channel flows (Zic et al, 2009). Its determination is an interdisciplinary task because according to Zic et al. (2009), it includes the knowledge of hydrology, statistics, hydromechanics, hydraulics, geology and mechanics. The value of n varies in space and time. It varies not only along the bed but also with the changes in water level. According to Abdul Ghaffar et al. (2004), most of the hydraulic computations related to indirect estimates of discharge require an evaluation of the roughness characteristics of the channel. Channel morphology depends on the interaction between fluid flow and the erodible materials in the channel boundary (Abdul-Ghafar et al, 2004). In general, effective factors in estimation of roughness coefficients include the roughness of channel bed materials and the cross section shape of the channel (Ebrahimi et al, 2008). Since the last century, many researchers in the developed countries have been developing different empirical equations for the determination of the roughness coefficient in order to adapt this to the natural morphology of the natural rivers in their respective countries. These researchers have used a number of flow resistance equations involving grain roughness, form roughness and a combination of both, but the Manning’s equation has been widely used internationally for predicting roughness values in channels (Abdul Gafar et al, 2004). One of the key steps in the process of mathematical modeling is to determine the relationship among the channel flow variables. This could be carried out through dimensional analysis; which is a method for Nigerian Institution of Agricultural Engineers © www.niae.net

112


helping determine how variables are related and for simplifying a mathematical model. It is based on two assumptions that: 1. Physical quantities have dimensions (Mass, M; Length, L; and Time, T) 2. Physical laws are unaltered when changing the units measuring the dimensions. It is a useful tool for developing prediction equation of various physical systems. The methodology of dimensional analysis reduces the physical quantities pertinent to a system to dimensionless groups (Degirmencioghi and Srivastava, 1996). It employs the Buckingham’s π- theorem to develop the relationship between the materials bed’s roughness, the water flow and the channel’s geometric parameters. The theorem states that: ‘If there are n variables (Independent and dependent variables) in a physical phenomenon and if these variables contain m fundamental dimensions (M, L, T), then the variables are arranged into (n-m) dimensionless terms. Each term is called π-term’. The equation is in the form:

( 1 ,  2 ,  3 ,......... .. m )  0

(1)

Dimensional analysis is a mathematical technique which allows the enlightened design of experimental investigation. It assists in experimental investigation by reducing the number of variables in the problem. The result of the analysis is to replace an unknown relation between n variables by a relationship between a smaller, n-m, of dimensionless groups. Any reduction in the number of variables greatly reduces the labour of experimental investigation. Many empirical formulae have been developed to estimate the values of roughness based on particle size distribution curve of surface bed material (French, 1985; Nguyen and Fenton, 2004). However, these formulae are often generated for natural rivers or in the laboratory set- up where all conditions are quite different from the field and not for constructed canals. The objective of this study, therefore, was to develop a roughness coefficient that would incorporate and relate the geometric and hydraulic properties of a channel with flow and fluid properties using dimensional analysis. 2.

MATERIAL AND METHODS

2.1

Experimental Site

The experiment was carried out at the National Centre for Agricultural Mechanization (NCAM), Ilorin. Ilorin is geographically located in the middle belt of Nigeria with a vegetation of derived savannah, and is situated on a longitude of 40 30’ E and latitude of 80 26’ N. It receives an average of 1200 mm annual rainfall. The soil of the experimental site is sandy loam and contains 12.48% clay, 18% silt and 69.52% sand. It is classified as Hyplustalf of Eruwa and Odo-owa series, developed from the parent materials consisting of micaceous schist and gneiss of basement complex which are rich in Ferro-magnesium materials (Ahaneku and Sangodoyin, 2003). 2.2

Experimental Design

Five different canal lining materials were examined. The five treatments were: Treatment I - Concrete (1:2:4) Treatment II – Clay- Cement (6:1) Treatment III - Burnt Cementitious Clay Treatment IV – Clay Soil Treatment V - Termite Mound.


113


Forty five (45) channels of these treatments were dug and laid in a completely randomized design, in a 5 3 x 3 factorial layout. The five treatments were replicated three times, using different levels of slopes for each material. The levels of slopes were 2%, 5% and 7%, respectively. Each channel was 15 m long, a length enough for the observation of all the flowing parameters, with side slopes of 2:1. 2.3


Water was discharged into the Channels corresponding to heads of 2 cm, 3 cm, 6 cm and 8 cm, respectively. Measurements of canal cross-section, including width of cross-section were taken for each of the canal and an automatic flow meter was used to determine the flow velocities at these heads. The wetted area, wetted perimeters and the hydraulic radii were determined at the various cross- sections. A relationship was obtained by dimensional analysis based on the Buckingham’s π theorem. The relationship obtained was subjected to regression analysis using the parameters/ data from the experimental set-up, viz: velocity (V), area (A), Depth of flow (d), hydraulic radius (R), discharge (Q) and slope (S). These data were subjected to multiple regression analysis. 2.4

Model Development

The major condition for the development of this model was that the flow in the channel was uniform. In the development of the equation, the roughness coefficient of the channel was considered to be a function of the hydraulic radius of the channel(R), the slope (S), the channel discharge, Q; flow depth, d; and acceleration due to gravity, g. This is expressed mathematically as in Equation 2. Table 1 is the summary of the variables and their dimensions. (2) n  f ( R, Q, S , , g, d , v) or (3) f (n, R, Q, S , , g , d , v)  0 The total number of variables, N is eight, while the number of fundamental dimension (m) is three, hence the number of π - terms is N-m i.e 8-3=5. It follows that the number of π - terms in the equation can be expressed as: (4) f ( 1 , 2, 3 4 , 5)  0

   

Therefore:

    

 R .  .V .d 1 a

b

c

(5)

 R .  .V .Q 2 a

c

(6)

 R .  .V .S a

3

b

b

c

(7)

 R .  .V .n 4 a

c

(8)

 R .  .V .g a

5

b

b

c

(9)

Where π1 to π5 are dimensionless terms, while a, b, and c are exponents to be determined by dimensional analysis. Table 1 Dimensions Influencing Roughness Coefficient Physical variables Symbols Hydraulic radius R Slope S Discharge Q Viscosity µ Acceleration due to gravity G Nigerian Institution of Agricultural Engineers © www.niae.net

Dimensions L L/L L3T-1 ML-1T-1 LT-2 114


Flow depth Velocity

D V

L LT-1

π1 - Term Replacing the right hand side of Equation 5 with the corresponding dimensions of the variables and the dimensionless term on the left hand side in the M0L0T0, Equation 10 was obtained as follows: M0L0T0= Lo (ML-1T -1)b (LT -1) (L

(10)

Equating the exponents of µ, L and T on both sides, we have Power of M, b = 0; Power of L, a – b + 1 = 0; Power of T, b – c = 0 By solving the above equation simultaneously, we have a = -1, b = 0, c= 0 Substituting the above values in equation (10) resulted to:



1

d R



(11)

Repeating the above procedure for the remaining dimensionless terms, Equations 12 to 15 were obtained:



  

2



Q

R

2

(12)

V

3

S

(13)

4

n

(14)

5



R

V

2

.g

(15)

Substituting the dimensionless terms 𝜋1, 𝜋2, 𝜋3 , 𝜋4, and 𝜋5 into Equation 4 yields the following expression:

f (n,

d Q R , 2 , S , 2 .g )  0 R RV V

(16)

Equation 16 shows the relationship between the various parameters and the different dimensionless combinations obtained. This expression does not give the exact relationship between these parameters. The relationship could only be determined through values from the experimental field data. Equation 4 can be written as a function of any of the others as in Equation 17:

 4  f ( 1 ,  2 ,  3 ,  5 )

d Q R n  f ( , 2 , S , 2 .g ) R RV V

(17) (18)

The dimensionless terms in Equation 17 are combined as in Equation 19.

 4  f ( 13 ,  25 )

(19)

The numbers of π terms are reduced by division (Equations 20 and 21) according to Shefii et al. (1996) as follows:

 13   1 1 x 3 

RS d


(20)

115


 25   2

1

R 2V Rg R 3 g x 5  x  Q V 2 QV

(21)

RS R 3 g 4  f ( , ) d QV

(22)

Therefore;

RS R 3 g n f( , ) d QV

(23)

a

 R 3 g   RS  n  K     QV   d 

b

(24)

The exponents of Equation 24 were determined by regression analysis using the experimental field data. The equation was first log - transformed to yield a linear expression for easy determination of the components as follows:

Y  ln K  ax1  bx2

(25)

R g ln n  ln K  a ln  QV  3

where:

  RS     b ln d    

(26)

Y  ln n  R3 g   X 1  ln QV    RS  X 2  ln   d 

Using data from the channel flow experiment, the values of a and b in Equation 24 were obtained by regression analysis using the SPSS 16.0 statistical package. The model coefficients are in Table 2, while the summary and the analysis of variance for the model are in Tables 3 and 4, respectively. The resulting regression equation is as follows:

 R3 g   RS    0.39 ln ln n  0.94  0.36 ln   d   QV 

(27)

From Table 2, ln K= -0.94; a = 0.36 and b= 0.39 Ln K= -0.94 Therefore, K = e-0.94 = 0.391 Coefficients a, b, and constant K from the regression equation were substituted in Equation 27 as exponents in Equation 28 as follows:


116


 R3 g  n  e  0.94   QV 

0.36

 RS   d   

0.39

(28)

n  R 3(0.36) g 0..36Q 0.36V 0.36 R 0.39 S 0.39 d 0.39 0.391

(29)

n  R1.08 (2.275)Q 0..36V 0..36 R 0.39 S 0.39 d 0.39 0.391

(30)

n  (2.275x0.391) R1.47 Q 0..36V 0..36 R 0..39 S 0.39 d 0.39

(31)

n  0.89R1.47 Q 0.36V 0.36 S 0.39 d 0.39

(32)

 R 1.47 S 0.39  n  0.89 0.36 0.36 0.39  Q V d 

(33)

Equation 33 can be used to predicte roughness coefficient. 2.5

Model Validation

The equation was validated using Manning’s roughness equation:

n

R

2

3

S V

1

2

(34)

The relationship between the predicted equation and the Manning’s equation is presented in Figure 1. The Manning’s and predicted roughness coefficients were compared as in Table 5 using the least significant difference (LSD); at 1% and 5% level of significance to ascertain whether there were significant differences between them. Table 2: Model Coefficients Unstandardized Coefficients

Constant (ln K) Ln R3g/QV Ln RS/d

B -0.94 0.36 0.39

Table 3: Model Summary R R Square 0.933 0.871

Std. Error 0.086 0.015 0.011

Adjusted R Square 0.870


Standardized Coefficients Beta 0.879 1.280

T -10.929 23.759 34.607

Sig. 0.001 0.001 0.001

Std. Error of the Estimate 0.062

117


Table 4: Analysis of Variance of the Model Sum of Square Df Regression 4.578 2 Residual 0.677 177 Total 5.255 179

Mean Square 2.289 0.004

F 598.83

Sig. 0.001

Table 5: Analysis of Variance between Manning’s and predicted Roughness Coefficient Sum of Squares Df Mean Square F Between Groups 2.7778E-07 1 2.7778E-07 0.072598ns Within Groups 3.3671E-04 88 3.8263E-06 Total 3.3699E-04 89 ns: No significant different at P≤0.05 3.


The regression analysis showed that there was high correlation between the dependent and independent variables. The relationship between the dependent variable and the independent variables is as in Figure 2. The coefficient of correlation from the regression analysis is 0.93 (Coefficient of Determination, R 2= 0.871). This shows that there was a very high correlation between the dependent and the independent variables. This implied that 87.1% relationship was found among the independent variables. The adjusted R2 showed that 87% of the variables can explain the model. Generally, the higher the value of the adjusted R2, the more significant the model. It is therefore, explicit to say that the variables can adequately explain the model. The validation of the predicted equation as shown in Figure 1 gave a very high correlation of 0.90 (Coefficient of Determination, R2 = 0.814). It also revealed that there was a good agreement between the predicted equation and Manning’s equation. The Manning’s and predicted roughness coefficients were compared in Table 5, using the least significant difference (LSD); at 1% and 5% level of significance; there were no statistical differences between them. Therefore, these results show that the interaction between the parameters of the channel has successfully been modelled using dimensional analysis.


118


0.032

np = 0.940n+ 0.001 R² = 0.814

0.030 0.028

np

0.026 0.024 0.022 0.020 0.020

0.025

n

0.030

0.035

Figure 1: Relationship between Manning and the Predicted Equation 0.035 0.030 0.025

Roughness Coefficient

0.020 n

0.015

np

0.010 0.005 0.000 1 3 5 7 9 111315171921232527293133353739414345

Replication Figure 2: Manning’s and predicted Roughness Coefficient

4.

CONCLUSIONS

A mathematical model was developed using dimensional analysis based on the Buckingham’s π theorem and a functional relationship between some channel and flow parameters was developed. The model was validated using Manning’s equation. The validation of the predicted equation gave a very high correlation of 0.90 with a Coefficient of Determination, R2, of 0.814. It also revealed that there was a good agreement between the predicted equation and Manning’s equation. Thus, the developed model in this study estimates the roughness coefficient with acceptable accuracy. Nigerian Institution of Agricultural Engineers © www.niae.net

119


REFERENCES Ahaneku, I.E. and A.Y. Sangodoyin. 2003. Parameter Estimates and Summary Statistics for TimeDependent Infiltration Models Under Varying Tillage Systems. Journal of Engineering and Engineering Technology. 3(2): 54-59. Degirmencioghi, A. And A.K. Srivastava. 1996. Development of Screw Conveyor Performance Models using Dimensional Analysis. Proceedings of the American Society of Agricultural Engineers. 39(5): 1757- 1763. Ebrahimi,N.G.,Fathi-Moghadam,M.,Kashefipour, S.M., Saneie, M. and K. Ebrahimi. 2008. Effect of Flow nad Vegetation States on River Roughness Co-efficients. Journal of Applied Sciences. 11: 2118-2123. French, R.H. 1985. Open Channel Hydraulics. McGraw- Hill, New York. Abdul-Ghafar, A.B., Ghani, A.A., Zakaria, N.A, Hasan, Z.A. and C.C. Kiat. 2004. Determination of Manning’s Flow Resistance Coefficient for Rivers in Malaysia. Proceedings of the International Conference on Managing Rivers in the 21st Century: Issues & Challenges. 1: 104-110. Nguyen, H.T. and J. D. Fenton. 2004. Using Two – Point Velocity Measurements to Estimate Roughness in Streams. Proceedings of the 4th Australian Stream Management Conference, Launceston, Tasmania, 19 – 22, October. Ed.I.D.Rutherfurd, Wiszniewski, M. Askey- Doran and R. Glzik. pp: 445 – 450. Dept. of Primary Industries, Water and Environment, Hobart. Shafii, S, S.K. Upadhyaya and A. R. E. Garett. 1996. The Importance of Experimental Design to the Development of Empirical Prediction Equations: A Case Study. Transaction of the American Society of Agricultural Engineers. 39(2): 377-384. Zic, E., Vranjes, M and Ozanic, N. 2009. Methods of Roughness Coefficient Determination in Natural River Beds. Proceedings of the International Symposium on Water Management and Hydraulic Engineering, Macedonia. Pp: 851-862


120


PHYSICO-CHEMICAL AND STRENGTH PROPERTIES OF SOME LOCAL MATERIALS FOR IRRIGATION CANAL LINING M. Y. Kasali1, A. O. Ogunlela2 and D. James1 and I. I. Azogu1 1 National Centre for Agricultural Mechanization (NCAM), Ilorin, Nigeria E- mail: [email protected] 2 Department of Agricultural & Biosystems Engineering, University of Ilorin, Ilorin, Nigeria E-mail: [email protected] ABSTRACT The physico - chemical and strength properties of some local materials for canal lining were investigated. These materials were: (i) Concrete (GC): which comprised of Cement, Sand and Granite of average size of 12 mm, in a ratio of 1:2:4. (ii) Termite Mound (TM) (iii) Clay Cement (CLC) (iv) Cementitious Clay (CCL), and (v) Clay Soil (CLS). Laboratory analysis of the chemical and physical properties of these materials was determined. The shear strengths and the cone indices were determined using the compaction characteristics of the materials. The moisture characteristics of the samples show that Concrete attained the maximum densities of 1.55 gcm-3, 1.57 gcm-3, 1.58 gcm-3 at the lowest moisture contents of 6.7%, 6.5 % and 7.0% at 5, 10, 25 blows, respectively, while the Clay - Cement attained maximum densities of 1.27 gcm-3, 1.30 gcm-3 and 1.33 gcm-3 at the highest moisture contents of 16.0%, 14% and 14.0% at the same levels of blows. The study showed that the plasticity of the materials increased as the clay content increased, while the shear strength and cone index increased with depth and decreased with increase in moisture content. KEYWORDS: Compactive effort, moisture content, dry density, shear strength, cone index, irrigation canal. 1.

INTRODUCTION

To provide irrigation water to crops, water has to be conveyed from the source to the field. Irrigation water conveyance had been an age long practice to get water conveyed from the source to the end users (Irrigation farmers). During this conveyance, considerable quantity of water is lost on transit. This loss has been a daunting problem facing local farmers because there is an irretrievable loss of valuable water resources. These losses in conveyance are majorly due to seepage and evaporation losses. Evaporation loss is a function of temperature, humidity and wind velocity. This type of loss is practically impossible to prevent while seepage losses can be prevented by the laying of impervious material along the channel. Most conventional methods used in preventing seepage losses are the use of compacted clay, tiles, soil- cement, concrete, etc. Seepage losses from irrigation channels have widely been identified as environmentally critical for the resulting groundwater accessions and associated drainage problems (Riaz and Sen, 2005). Seepage, therefore, has a very adverse effect on the surrounding of the canal. It often creates a localized high water table that damages crops in adjacent fields due to waterlogging and soil salinization. The need for the improvement of the engineering properties of soil has been an age long practice. When soil is to be constructed upon, sometimes the soil at the site may manifest weak properties and the soil is removed and replaced with stronger soil but this method is cumbersome and is being replaced by modern technique that involves the improvement of the engineering properties of the site soil. Local materials, if carefully prepared, will help to control excessive water losses in irrigation water conveyance as the conventional lining materials. Since concrete, which has been the conventionally used Nigerian Institution of Agricultural Engineers © www.niae.net

121


lining material is becoming expensive; there is the need to search for alternative local materials that are in sufficient quantity in farmers’ vicinities. These materials must be able to replace concrete in terms of seepage reduction and durability. The objective of this study was therefore to investigate the strength of these local lining materials to ascertain their potentials and suitability for irrigation canal lining. 2.


2.1

Description of the Study Area

The experiment was carried out at the National Centre for Agricultural Mechanization (NCAM), Ilorin. Ilorin is geographically located in the middle belt of Nigeria with a vegetation of derived savannah, and is situated on a longitude of 40 30’ E and latitude of 80 26’ N. It receives an average of 1200 mm annual rainfall. The soil of the experimental site is sandy loam and contains 12.48% clay, 18% silt and 69.52% sand. It is classified as Hyplustalf of Eruwa and Odo-owa series, developed from the parent materials consisting of micaceous schist and gneiss of basement complex which are rich in Ferro-magnesium materials (Ahaneku and Sangodoyin, 2003). 2.2

Materials

Five sample materials were considered for study. These materials were: (i) Concrete (GC): which comprised of Cement, Sand and Granite of average size of 12 mm, in a ratio of 1:2:4. (ii) Termite Mound (TM) (iii) Clay - Cement (CLC) (iv) Cementitious Clay (CCL), and (v) Clay Soil (CLS). 2.3

Determination of Particle Size Distribution and Chemical Composition

Samples of each of the treatments were collected for particle size distribution analysis and texture. The soil samples were air dried and passed through a 2-mm sieve to remove stones and crumbs. The particle size distribution was obtained through sieve analysis of the grains of the samples to determine the sand fraction. The known weight of each of the samples was allowed to pass through standard set of sieves and the weight of the fractions retained on each sieve is recorded. These weights were expressed as the percentages of the total weight of the samples. The exchangeable Magnesium was extracted and titrated with sulphuric acid, while available phosphorous and potassium were extracted using double acid solution of 0.05N hydrochloric acid and 0.025N sulphuric acid. Sodium was also extracted and titrated with sulphuric acid. Calcium and Magnesium were determined using absorption spectrophotometer. The organic matter contents of the samples were estimated from the carbon content of the sample using the method of Walkley and Black (1934). 2.4

Consistency Limits and Hydraulic Conductivity

The Atterberg limits (plastic and liquid limits) were determined using Cassagrande method as described by Arku and Ohu (1991). The Plasticity Index (PI) was determined as in Equation 1: PI = WL - WP

1

where: WL = liquid limit; WP = plasticity limit; PI = Plasticity Index The permeability (saturated hydraulic conductivity) of each sample was determined using the falling head permeameter. Nigerian Institution of Agricultural Engineers © www.niae.net

122


The hydraulic conductivity was determined as in Equation 2:

K

QL Ah

2

where: K = Hydraulic conductivity, cm/s; L = Sample length, cm; A = Area of sample, cm2 2.5

Compaction Characteristics

The compaction characteristics were determined using the standard compaction mound. The samples were subjected to 5, 15 and 25 blows of a standard proctor hammer of 2.5 kg in cylindrical mould of 105 mm diameter and 115 mm height, at different moisture contents following the proctor compaction procedure (Lambe, 1951). The dry densities were determined at four repeated times for each sample at every moisture content and compactive effort. 2.6

Cone Index and Shear Strength of Compacted Samples

The shear strength determination of compacted samples was conducted by sieving the samples through 2 mm sieve size and then compacted. The cone indices of each sample were measured using the Farnell hand held soil penetrometer, fitted with a 300 cone. The resulted graphs of the cone indices against the corresponding depths are as shown in Figures 6 – 20. The shear strengths were also determined for each sample using the Pilcon hand held vane tester with 33 mm vanes. The vane was pushed vertically into the compacted soil to different depths of 5, 15 and 25 levels of blows, respectively. At the end of the shearing, the vane returns almost instantaneously in the anticlockwise direction and the shear strengths were read from the dial of the Vane tester. 3.


3.1

Particle Size Distribution, Chemical Composition, Consistency Limits and Hydraulic Conductivity

The textural classes and chemical compositions of the samples are show in Tables 1 and 2. From the grain size analysis, it was found that the grain seizes of the five samples were distributed within the following ranges; 6-38% silt, 8.48-38.43 clay and 43.57-82.52 % sand. The textural classifications and the chemical composition of the samples are in Tables 1 and 2, respectively. The liquid limit, plastic limit and the plasticity index values representing the soil types were found to be in the range of 34-49%, 17-24.3% and 17-24.7%, respectively (Table 3). Table 3 shows that the samples have average values of liquid limits and plasticity index. Clay Cement mixture has the highest plasticity index of 24.7%, while, Concrete has the lowest of 17%. Termite mound, Cementitious Clay and Clay Soil samples have 19.2%, 19.5% and 19.6%, respectively. The soils were classified into inorganic clays of medium plasticity according to the Cassagrande plasticity chart in the Unified Soil Classification System (USCS) in compliance with ASTM standards (ASTM D 2487-00 standard) as employed by Ince and Ozdemir (2003). This shows that all the materials are workable and are capable of carrying considerable loads. Generally, conductivity is affected by the size and distribution of soil particles which generally influence the size of voids conducting flow. The factors that affect hydraulic conductivity are mineral composition, texture, particle size distribution, characteristics of wetting fluid, exchangeable- cation, void ratio and Nigerian Institution of Agricultural Engineers © www.niae.net

123


degree of saturation of the medium. Results from Table 3 shows that all the samples have medium permeability and could be good materials for canal lining, if properly compacted. The plasticity indices of the clay - cement and clay soil were the highest of the samples which might be due to the fact that they contain higher silt and lower sand percentage than other samples. It could be observed in Table 1 that the clay and silt contents of the samples decreased as the sand content increased. Similarly, increase in plasticity index with an increase in clay content was observed. This trend in results was in conformity with the results obtained by Ekwue et. al. (2002) and Adekalu et. al. (2007). Table 1. Textural and Organic Properties of the Samples Samples+ CCL

Components (%) Organic Carbon

GC 0.02

TM 0.51

0.24

4.76

2.15

Organic Matter

0.05

0.87

0.67

8.22

3.71

Sand

82.52

59.52

47.52

53.52

43.57

Silt

6.0

30.0

20.0

38.0

18.0

Clay

11.48

10.48

32.48

8.48

38.43

CLS

CLC

+

GC= Concrete TM= Termite Mound CLC= Clay- Cement CCL= Cementitious Clay CLS= Clay Soil Table 2. Chemical Properties of Samples Components GC TM CLC N(%) 0.003 0.07 0.028 Ca2+ (mg/Kg) 32.52 67.52 16.43 Mg2+ (mg/Kg) 2.58 31.17 1.60 + (mg/Kg) Na 0.049 125.11 129 (mg/Kg) P 0.124 120.54 203.25 Ph (mg/Kg) 34.0 33.97 27.55 Cl2- (mg/Kg) 0.027 20.38 12.65 (mg/Kg) Co3 2.81 8.81 12.93 Si 4.38 6.62 GC= Concrete TM= Termite Mound CLC= Clay- Cement CCL= Cementitious Clay CLS= Clay Soil

Samples CCL 0.6 77.9 41.56 142.0 154.78 58.05 29.26 46.90 -

CLS 0.09 36.36 20.78 136.42 133.36 34.97 32.07 21.45 -

Table 3. Physical and Index Properties of the Samples Properties Samples+ GC TM Bulk Density (kg/m3) 1.50 1.49 Dry Density (kg/m3) 1.45 1.47

CLC 1.50 1.45

CCL 1.47 1.43

CLS 1.57 1.49

Specific Gravity Liquid Limit (%) Plastic Limit (%) Plasticity Index (%) Permeability (cm/sec)

2.60 49.0 24.3 24.7 5.63 x

2.67 41.0 21.5 19.5 1.07 x

2.63 37.0 17.4 19.6 8.65 x

2.68 34.0 17.0 17.0 8.57 x

2.65 39.0 19.8 19.2 2.55 x


124


10-5 10-4 10-5 GC= Concrete TM= Termite Mound CLC= Clay- Cement CCL= Cementitious Clay CLS= Clay Soil

10-5

10-5

+

From the results of the dry density- moisture relationships (Table 4), it was observed that Concrete reached the maximum dry density at the lowest moisture contents. This corroborates and supported the fact that the sample has higher percentage of sand than the other samples and the plasticity of the sample was the lowest among the samples as revealed in Table 1. The reverse was the case for Clay- Cement, which had the highest plasticity index of 24.7 % and the lowest percentage of sand. This shows that the value of plasticity index of a soil reflects the clay contents in the soil and a high value will reflects the workability of the sample due to cohesion between the soil’s grain particles. The compaction tests revealed that the dry densities of the samples increased with compactive efforts, which shows that dry density is a function of moisture content and compactive effort. The peak of each curve shows the maximum dry density for a given compactive effort. The results of the compaction test as revealed in Table 4 and Figures 1- 5, could be explained by the fact that at the side of the optimum water content, the dry density increases with the increasing water content. This is probably due to the development of large water film around the particles, which tends to lubricate the particles and makes them easier to be moved about and re-orientate into a denser configuration (Holtz and Kovacs, 1981). At the wet side of the Optimum Moisture Content (OMC), water starts to replace soil particles in the compaction mould and since the units weight of water is much less than the unit weight of soil, dry density decreases with the increasing water content. The table shows that the maximum dry densities of 1.55gcm-3, 1.57 gcm-3, 1.58 gcm-3 were attained by concrete sample at 5, 10 and 25 blows, at the lowest level of moistures of 6.7%, 6.5 % and 7.0%, respectively. This was followed by Termite Mound sample with maximum dry densities of 1.45 gcm-3, 1.51gcm-3, and 1.63 gcm-3 at moisture levels of 10.4%, 10.1 % and 9.0%, respectively. Clay soil sample had maximum dry densities of 1.5 gcm-3, 1.57 gcm-3 and 1.56 gcm-3 at moistures of 11.6 %, 11.1 % and 10.1 %, respectively. This was followed by Cementitious clay samples with densities of 1.34 gcm-3, 1.38 gcm-3 and 1.44 gcm-3 at moisture of 14.0 %, 15.2 % and 13.5 %, respectively, while the Clay cement sample had the least densities of 1.27 gcm-3, 1.30 gcm-3 and 1.33 gcm-3, respectively. Results further revealed that an increase in compactive effort increases the maximum dry density but decreases the optimum water content. This is because higher compactive effort yields more parallel orientation of the clay particles, which allow for closer particle orientation and hence a higher unit weight of soil (Holz and Kowacs,1981; Ige and Ogunsanwo,2009). This was manifested in all the samples and it shows that at a higher compactive effort, the grain particles of the soil become close together and the unit weights of the samples increase. These results conform with the results obtained by Ige and Ogunsanwo (2009). The implication of this is that channels with adequate compaction will reduce hydraulic conductivity and hence drastic reduction in seepage.


125


Table 4. Summary of Optimum Moisture Contents and Dry Densities of Samples Treatments Optimum Moisture Content (%) Dry Density( kg/m3) Number of Blows 5 15 25 5 15 10.4 10.1 9 1.45 1.51 TM 11.6 11.1 10.1 1.45 1.57 CLS 14.0 15.2 13.5 1.34 1.38 CCL 16.0 14.0 14.0 1.27 1.30 CLC 6.7 6.5 7.0 1.55 1.57 GC GC= Concrete TM= Termite Mound CLC= Clay- Cement CCL= Cementitious Clay CLS= Clay Soil

1.7

25 1.63 1.56 1.35 1.33 1.58

1.6 5 blows

Dry Density, kg/cm3

Dry Density, kg/cm3

5 blows 1.6

1.5

1.5

1.4

1.4

1.3

1.3

1.2 0

5

10

15

20

0

5

Fig.1.Effect of Moisture Content on Dry Density of Termite Mound

1.3 1.2 1.1 5

10

15

20

Moisture Content, %

Fig. 3. Effect of Moisture Content on Dry Density of Cementitious Clay

1.6

20

5… 15…

1.3 1.25 1.2 -5

5 15 Moisture Content, %

25

Fig. 4. Effect of Moisture Content on Dry Density of Clay Cement

5 blows

1.58

Dry Density, kg/cm3

Dry Density, kg/cm3

Dry Density, kg/cm3

1.35

5 blows 15…

0

15

Fig. 2.ffect of Moisture Content on Dry Density of Clay Soil

1.5 1.4

10

Moisture Content, %

Moisture Content, %

1.56 1.54 1.52 1.5 0

Moisture 5 Content, %

10

Fig. 5. Effect of Moisture Content on Dry Density of Granite Cement


126


3.3

Cone Index and Shear Strength

The effects of moisture content on cone index and shear strength at 5, 15 and 25 blows are as in Figures 6 - 20 and 21-35, respectively. The cone indices and the shear strengths of the soils increase with depth and compaction level and as the moisture contents decrease. This trend conforms with the works of Al Rawas (2006); Adeniran and Babatunde (2010) and Manuwa, (2009). This is expected because as the moisture increases, the soil strength decreases and hence the resistance to penetration decreases. The highest cone index of 285.0 kPa was obtained in Clay - Cement sample, while the lowest of 55 kPa was from Concrete sample. This might be due to large voids between the granite particles and the cement, which allows for easy penetration of the Concrete mix and also the higher percentage of clay in the clay – cement, but as the moisture reduces in the Concrete composite, the penetration resistance is expected to increase because the voids between the granite and the sand grains would be closed up through the reaction between water, sand particles and cement (hydration) to form a very hard aggregate. Clay had the highest shear strength of 126 kPa, while Concrete had the lowest of 28 kPa. The same reasons for the cone index could also be deduced for this trend.


127


40 6

9

12

120 80 40 0

15

9 12 Moisture Content, %

3 cm depth 6 cm depth 9 cm depth 12 cm depth

120 80 40 0 6

9

12

15

18

160

80 40 0

0

5

10

15

20

25

Moisture Content, %

Cone Index, kPa

Fig.15. Effect of Moisture Content on Cone Index of Clay Cement Composite at 5 Level of Blows

66 60 54 48 42 36 30 24 18 12 6 0

At 3 cm At 6 cm At 9 cm At 12 cm

0

3

6

9

12

Moisture Content, % Fig. 18. Effect of Moisture Content on Cone Index of Granite Cement Mxture at 5 Level of Blows

10

12.5

15


160 120 80 40

3

300 280 260 240 220 200 180 160 140 120 100 80 60 40 20 0

5 10 15 20 Moisture Content, %

200 120 80 40 0

3

6

9

12

15

18

300 270 240 210 180 150 120 90 60 30 0

21

12

Fig. 19. Effect of Moisture Content on Cone Index of Granite Cement Mixture at 15 Level of blows


0 5 10 15 20 Moisture Content, %

25

Moisture Content, %


9

Moisture Content, %


0


160

Fig. 14. Effect of Moisture Content on Cone Index of Cementitious Clay at 25 Level of Blows

Fig. 16. Effect of Moisture Content on Cone Index of Clay Cement Composite at 15 Level of Blows

78 72 66 60 54 48 42 36 30 24 18 12 6 0

15

Fig. 11. Effect of Moisture Content on Cone Index of Clay at 25 Level of Blows


0

6 9 12 Moisture Content, %

6

Fig. 13. Effect of Moisture Content on Cone Index of Cementitious at 15 Level of Blows

Cone Index, kPa

At 3 cm At 6 cm At 9 cm

7.5

Moisture Content, % Fig. 8. Effect of Moisture Content on Cone Index of Termite Mound at 25 Level of Blows

6 9 12 15 18 21 Moisture Content, %

21

Cone Index, kPa

Cone Index, kPa

220 200 180 160 140 120 100 80 60 40 20 0


120

Moisture Content, % Fig. 12. Effect of Moisture Content on Cone Index of Cementious Clay at 5 Level of Blows

40

15

Fig. 10. Effect of Moisture Content on Cone Index of Clay at 15 Level of Blows

Cone Index, kPa

Cone Index, kPa

160

80

0 6

Moisture Content, % Fig. 9. Effect of Moisture Content on Cone Index of Clay at 5 Level of Blows

120

200


160

0

160

5

Cone Index, kPa

80

Cone Index, kPa

9 12 15 Moisture Content, % Fig. 7. Effect of Moisture Content on Cone Index of Termite Mound at 15 Level of Blows

Cone Index, kPa

Cone Index, kPa

120

6

200


160


200

0

Cone Index, kPa

3 6 9 12 15 Moisture Content, % Fig. 6. Effect of Moisture Content on Cone Index of Termite Mound at 5 Level of Blows

240


Cone Index, kPa

0

200

200 180 160 140 120 100 80 60 40 20 0

25

Fig. 17. Effec of Moisture Content on Cone Index of Clay Cement Composite at 25 Level of Blows

Cone Index, kPa

3 cm depth 6cm depth 9 cm depth 12 cm depth

Cone Index, kPa

Cone Index, kPa

200 180 160 140 120 100 80 60 40 20 0

78 72 66 60 54 48 42 36 30 24 18 12 6 0


0

3

6

9

12

Moisture Content, % Fig. 20. Effect of Moisture Content on Cone Index of Granite Cemente Mixture at 25 Level of blows

128

Journal of Agricultural Engineering and Technology (JAET), Volume 22 (No.2) June, 2014 120 100

40 20 0 9

12

15

60 40 20 6

9

Moisture Content, %

Shear Strength , kPa

40 20 0


100 80 60 40 20 0

6

9

12

15

6

9

80

40 20 0 12

15

18

21

Moisture Content, %

Shear Strength, kPa

Fig. 27. Effect of moisture content on shear strength of cementitious clay at 5 level of blows

80

40 20 0

0

5

10

15

20

Moisture Content, %

Shear Strength, kPa


0

3

6

9


3

6

12

15

120 100 80 60 40 20 0

18

12

Moisture Content, % Fig. 33. Effect of moisture content on Shear Strength of Granite Cement mixture at 5 Level of blows

5

10

15


60 40 20 0 6

20

25

24 18 12 6 0

0

3

6

9

12



12

15

18

21

160 140 120 100 80 60 40 20 0


0

5

10

15

20

25

Moisture Content, %


30

9

Fig. 29. Effect of Moisture Content on Shear Strength of Cementitious Clay at 25 Level of Blows

Moisture Content, %

36

15

Moisture Content, %

Fig.31. Effect of Moisture Content on Shear Strength of Clay Cement Composite at 15 Level of Blows 42

12

80

21


0

9

100

Moisture Content, %

25

Fig.30. Effect of Moisture Content on Shear Strength of Clay Cement Composite at 5 Level of Blows

30 27 24 21 18 15 12 9 6 3 0

9

Fig. 28. Effect of Moisture Content on Shear Strength of Cementitious Clay at 15 Level of Blows


60

0 6

15

Moisture Content, %

20

Shear Strength, kPa

9

12

Fig. 26. Effect of Moisture Content on Shear Strength of Clay on 25 Level of Blows

40

Shear Strength, kPa

6

15


60

Shear Strength, kPa

Shear Strength, kPa

60

12

Fig. 25. Effect of Moisture Content on Shear Strength of Clay on 15 Level of Blows


9

140 120 100 80 60 40 20 0

Moisture Content, %

Moisture Content, % Fig. 24. Effect of Moisture Content on Shear Strength of Clay at 5 Level of Blows

80

6

Fig. 23. Effect of Moisture Content on Shear Strength of Termite Mound at 25 Level of Blows

Shear Strength, kPa


60

120

3

Moisture Content, %

Moisture Content, %


80


15



100

12


6

80

140 120 100 80 60 40 20 0

Shear Strength, kPa


60


80




Fig.32. Effect of Moisture Content on Shear Strength of Clay Cement Composite at 25 Level of Blows

Shear Strength, kPa

100

42 36 30 24 18 12 6 0


0

3

6

9

12


129


4.

CONCLUSIONS

These local materials are very promising for utilization in irrigation canal lining in terms of strength and seepage reduction as revealed by their strengths and hydraulic conductivities. Their performances vis-avis concrete is very encouraging, and could perform well if employed in small irrigation canal linings but may not compete in terms of durability. It can therefore be concluded that linings made of local materials have the potential of reducing seepage on a permanent basis, though not as satisfactory as that of concrete but the need for economy and homeward exploitation of these materials that are indigenous and available in the farmers’ environs has supported their utilization. REFERENCES Adekalu,K.O.; D.A. Okuade and J.A. Osunbitan. 2007. Estimating Trafficability of Three Nigerian Agricultural Soils from Shear Strength – Density – Moisture Relations. Journal of International Agrophysis. 21:1-5. Adeniran, K.A. and O. O. Babatunde. 2010. Investigation of Wetland Soil Properties affecting Optimum Soil Cultivation. Journal of Engineering Science and Technology Review. 3(1): 23 – 26. Ahaneku, I.E. and A.Y. Sangodoyin. 2003. Parameter Estimates and Summary Statistics for TimeDependent Infiltration Models Under Varying Tillage Systems. Journal of Engineering and Engineering Technology. 3(2): 54-59. ASTM D 2487-00,2003. Standard Practice for Classification of Soils for Engineering purposes (Unified Soil Classification System). In: Annual Book of ASTM Standards. 4:248-259.West Conshohocken, PA. Al-Rawas, A. A. and A. W. Hago. 2006. Evaluation of Field and Laboratory Produced Burnt Clay Pozzolans. Journal of Applied Clay Science. Vol. 31, pp: 29-35. Arku, A.Y. and J.O. Ohu. 1991. Optimum Moisture Content for Compacting Agricultural Soils in Borno State. A paper presented at the 15th Annual Conference of the NSAE, held at the University of Maiduguri, 10th – 13th September. Ekwe, E.I, Stone, R.J. and Ramphalie, S. (2002). Engineering Properties of some Wetland Soils in Trinidad. Appllied Engineering in Agriculture. 18(1):37 – 45. Holtz, R.D. and W.D. Kovacs. 1981. An Introduction to Geotechnical Engineering. Prentice- Hall Inc., Englewood, Cliff, N.J. Pp: 109-165. Ince,I and Ozdemir. 2010. Ozean Journal of Applied Sciences 3(3), pp: 357-362. Lambe, T. W. 1951. Soil Testing for Engineering. John Wiley Press, New York. Manuwa,S.I. 2009. Properties and Shear Strength of Insitu and Compacted Termite Mound Soil (TMS) in Akure South–Western Nigeria. Proceedings of the 18th ISTRO Conference Held in Izmir, Turkey between June 15 &19. Riaz, M and Z. Sen. 2005. Aspect of Design and Benefits of Alternative Lining Systems. Journal of European Water 11(12): 17-27. Ige, O.O. and O. Ogunsanwo. 2009. Assessment of Granite- derived Residual Soil as Mineral Seal in Sanitary Landfills. http:// www. Sciencepub.net/ researcher. Pp;80-86. March 27, 2013. Walkle, A and I.A. Black. 1934. An Examination of Dedjareff Method for Determining Soil Organic Matter and Proposed Modification of the Chronic Acid Titration Method. Soil Science. S37: 29-38.


130


WATER TEMPERATURE SIMULATION OF OWENA RIVER AT OWENA DAM SITE USING HEC-RAS MODEL G.I. Ezenne, E.L. Ndulue and O.L. Yakubu Department of Agricultural & Bioresources Engineering, University of Nigeria, Nsukka, Nigeria E-mail: [email protected] ABSTRACT Water temperature happens to be one of the most important physical parameter used in describing water quality. Temperature affects almost all processes and other water quality parameters. The processes include stratification, biotic processes and chemical fate processes. Water temperature affects other water quality parameters such as Dissolved Oxygen and Biochemical Oxygen demand (BOD). Today, with recent technological advances, several computer-based hydrologic/water quality models have been developed for applications in hydrologic modelling and water resources studies. The HEC-RAS (Hydrologic Engineering Centers Rivers Analysis) model performs one-dimensional steady flow, unsteady flow, sediment transport/mobile bed computations, and water temperature modeling. This study tries to simulate the water temperature of Owena dam using HEC-RAS model. The input data for the simulation are geometric data, steady flow data, and meteorological data. Input data was entered in the various data editor windows and interfaces of the model in a logical sequence using a series of components parts of the model that direct the program to accept, process, compute and store various parameters and their corresponding values. The results obtained were compared with the observed temperature data of Owena dam for the months of January, June and July. The result of the observed and simulated water temperature shows a correlation of determination (R2) of 0.92, 0.83 and 0.6 for the months of January, June and July respectively. From the result, it can be concluded that HEC-RAS can actually simulate water temperature of Owena River in Ondo state of Nigeria. 1.

INTRODUCTION

Water quality is defined as the physical, chemical and biological characteristics of water. Water quality determines the ‘goodness’ of water for particular purposes. Water quality parameters include temperature, dissolved oxygen, nutrients (phosphates and nitrates), pH, turbidity, biochemical oxygen demand (BOD), alkalinity, acidity, carbon dioxide, and specific conductance. Water quality is greatly influenced by the geological structure and the lithology of the watershed or aquifer, the chemical reactions that takes within the watershed or aquifer, as well as the type of land uses, anthropogenic activities and the climatic characteristics of the watershed or basin (Tsakiris and Alexakis, 2012). One of the man-made problems associated with water quality is thermal pollution. Thermal pollution is the introduction of warm water or other substrates into an aquatic ecosystem. Sources include industries such as power plants and also storm-drain runoff which has been warmed on streets, parking lots and sidewalks. In addition, human activities such as construction of roads, deforestation and the removal of vegetation around the water can lead to an increase in water temperature. Of all the above listed water quality parameters, water temperature is the easiest and often least costly to monitor in field conditions. It is usually measured by thermometer, thermocouple or thermostat. Water temperature is one of the most important physical characteristics of water. It is an important property that determines water suitability for human use, industrial applications and aquatic ecosystem functioning. Temperature impacts both the chemical and biological characteristics of surface water. Park and Clough (2012), described temperatures as driving variable which force systems behave in certain ways. Temperature affects both the processes in the aquatic system and other water parameter. Virtually all processes are temperature-dependent. They include stratification; biotic processes such as decomposition, photosynthesis, consumption, respiration, reproduction, and mortality; and chemical


131


fate processes such as microbial degradation, volatilization, hydrolysis, and bioaccumulation(Park and Clough, 2012). Temperature exerts a major influence on aquatic organisms with respect to selection/occurrence and level of activity of the organisms. Increase in temperature also cause changes in aquatic plants and animals. As the temperature increases, the rate of photosynthesis increases. At temperatures above 32oC, the rate of photosynthesis will start to level off and then begin to decrease as the temperature continues to increase. As photosynthesis increase, the number of aquatic plants increases. This can lead to an increase in the number of plants which die and are decomposed by aerobic bacteria which consume oxygen in the process. Long-term shifts in temperature can also result in a change in the composition of organisms that make a stream their home. In general, increasing water temperature results in greater biological activity and more rapid growth. Temperature also affects the metabolic rate of aquatic animals, rates of development, timing and success of reproduction, mobility, migration patterns and the sensitivity of organisms to toxins, parasites and disease. Life cycles of aquatic organisms are often related to changes in temperature (Stephanie, 2008). Temperature also affects other water quality parameter like Dissolved Oxygen (DO), an important water characteristic that strongly affects many aquatic organisms. The solubility of oxygen increases as temperature decrease. This means that cold water holds more oxygen than hot water. The reason for this is, when water boils, water is converted to water vapour. The water vapour released contains dissolved oxygen. The amount of oxygen that will dissolve in water increases as temperature decreases. Water at 0oC will hold up to 14.6 mg of oxygen per litre, while at 30oC it will hold only up to 7.6 mg/L. Generally, an increase in temperature will decrease the solubility of oxygen. Temperature also affects the biochemical oxygen demand. An increase in temperature brings about a decrease in dissolved oxygen, which will increase the metabolic rate, thereby increasing biochemical oxygen demand (BOD). With recent technological advances, several computer-based hydrologic/water quality models have been developed for applications in water resources studies. The need for more scientifically sound analyses has led to the creation of a large number of water quality models. Technologically-based tools such as models and geographic information systems (GISs) can provide increased clarity on probable or alternative outcomes, and thus enable decision-makers to more effectively use traditional planning tools. Models are described as an approximate description of a class of real-world objects and phenomena expressed by mathematical symbolisms (Agoshkov, 2002). A detailed list of Hydrological and water quality models is given by (Moriasi et al., 2012; Berit and Jonas, 2004). With the predictive ability of models, they serve as management tool for decision makers. Examples include HSPF (Bicknell et al., 2001), QUAL2K (Chapra et al., 2008), WASP (Ambrose et al., 1993), AQUATOX (Park and Clough, 2012), SPARROW (Schwarz et al., 2006), SWAT (Arnold et al., 1993), HEC-RAS (HEC-RAS, 2010) to mention a few. The modeling results from these models under different pollution scenarios are very important components of environmental impact assessment and can provide a basis and technique support for environmental management agencies to make right decisions (Qinggai et al., 2013). Water quality studies carried out using models include the following: Vasudevan, M. et al. (2011) applied the QUAL2K model to access waste loading scenario in River Yamuna; Picket (1997) used the WASP5 for Total Maximum Daily Load (TMDL); for the Black River in Washington State, US. (Love and Donigian, 2002) applied the HSPF model for nutrient loadings on Long Island Sound watersheds in Connecticut; Abbaspour et al. (2007) applied the SWAT to predict nitrate and total phosphorus on the 1700 km2 Thur River basin in Switzerland; Gassman et al. (2007) gave a detailed application of the SWAT model. The HEC-RAS model performs one-dimensional steady flow, unsteady flow, sediment transport/mobile bed computations, and water temperature modeling. This component of the modeling system is intended to allow the user to perform riverine water quality analyses. An advection-dispersion module is included with this version of HEC–RAS, adding the capability to model water temperature. This new module uses the quickest-ultimate explicit numerical scheme to solve the 1-Dimensional advection-dispersion equation using a control volume approach with a fully implemented heat energy budget. Transport and Fate of a limited set of water quality constituents is now also available in HEC-RAS. The currently available water Nigerian Institution of Agricultural Engineers © www.niae.net

132


quality constituents are: Dissolved Nitrogen (NO3-N, NO2-N, NH4-N, and Org-N); Dissolved Phosphorus (PO4-P and Org-P); Algae; Dissolved Oxygen (DO); and Carbonaceous Biological Oxygen Demand (CBOD). Amod et al., (2012) applied the HEC-RAS model in hydrologic and water temperature modeling in Alameda creek. Thus this study was undertaken to simulate water temperature of Owena River at Owena Dam site using the HEC-RAS. 2.

METHODOLOGY

2.1

Study Area

The study area is located in Ondo State of Nigeria, and lies between the latitudes 7 0 35’ and 70 00’ and longitudes 40 50’ and 50 15’. The dam axis lies about 17km north of the point where the Owena River crosses the Ondo-Akure road and is located on latitude 70 18’ and longitude 50 00’. The Owena Multipurpose Dam is an earth dam with a total fill volume of 1,014,729m3 and a height of 24.2m above river bed, the dam crest level being at elevation 313.2m. It is 1457m long along the crest and it is designed to impound 30 million cubic metres of water in its reservoir that has a total surface area of 7.4 square kilometres. The dam axis cuts across a valley which has a fairly symmetrical cross-section, with an average transverse slope of approximately 6.5% on the left bank and 5.0% on the right bank. Cocoa plantations cover most of the right bank area and a forest reserve is located on the left bank. The river flows mainly in the North-South direction. Its width is in the range of 10-15m; its banks are steep and about 2m in height on either side. The site lies within the South Western basement area of Nigeria in which the bedrock is essentially of the basement complex series. The basement out-crops at some places, especially along the river bed. Outcrops equally abound at both the abutments as well as within the reservoir area. The catchment area of the dam is 738 km2. Figure 1 shows the map of the study area.

SCALE 1:2000 Figure 1: Map of the study area showing features around it. Nigerian Institution of Agricultural Engineers © www.niae.net

133


2.2

Data Requirements and Organization

The water quality data entry menu, which manages input data, calibration of parameters, and finally output tools menu which manage output files to facilitate viewing and exporting model results was used. Water quality boundary data, meteorological data (atmospheric pressure, air temperature, humidity (vapor pressure, relative humidity, wet bulb or dew point), solar radiation, wind speed, and cloudiness) and source and sink parameters were entered in the water quality data window. This window was accessed from the main water quality input either through the menu bar by selecting EDIT- water quality data or by selecting the water quality data icon. The water quality data entry window opened when Edit was selected or alternatively by selecting the water quality data icon. This water quality data entry window is divided into three panes. The navigation bar is oriented as a vertical column at the far left. Its tree structure allows the user to access all input data and parameters. The two panes to the right of the navigation bar change in response to the selection on the left. To start a new water quality analysis, top row of the Navigation Bar (the line that says New Water Quality File) was selected, and a name for the data set was entered. Next, in the constituent selection panel, temperature modeling was then turned on to start the modeling. Entering and editing table data and meteorological data involves the following steps: time series generation tool, entering initial conditions, entering an initial distribution, entering dispersion coefficients and entering meteorological data. In order to model water temperature, at least one full meteorological data set must be available. The model supports multiple meteorological data sets. Each water cell was individually assigned to a particular data set. The entry of meteorological data set for the simulation followed a sequential order as arranged below: atmospheric pressure; air temperature; humidity (vapor pressure, relative humidity, wet bulb or dew point); solar radiation; wind speed; and cloudiness. A time series of air temperature, humidity, and wind speed radiation with a sampling frequency of at least once per three hours was necessary for simulation of water temperature variation. In addition to meteorological time series, each data set includes a limited amount of physical information including latitude, longitude, and site elevation. 3.


Table 1 shows results of simulated water temperature at Owena dam for the months of January, June and July respectively. Simulated water temperature obtained using HEC-RAS was compared with the observed water temperature data of Owena dam. The time series plots for simulated and observed water temperature of Owena dam for the months of January, June and July are shown in Figures 2-4. Table 1: Simulated water temperature at Owena dam for the months of January, June and July Date Time(Hours) Simulated Water Temperature (0C) at Owena Dam 1 2 3 4 5 6 7 8

09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00

January, 2011

June, 2011

July, 2011

17.20 18.99 19.38 21.17 22.34 22.10 22.18 22.49

19.20 20.06 18.34 21.24 19.28 18.73 16.30 20.61

19.20 19.03 19.70 18.70 17.70 19.53 20.11 17.87


134


9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00 09:00:00

23.27 23.35 23.51 23.89 23.51 24.05 24.83 25.68 25.84 27.09 26.07 27.32 27.48 27.16 26.85 25.61 26.54 26.62 26.39 27.32 27.48 26.85 26.23

17.08 18.10 16.06 14.41 16.53 17.86 20.06 18.89 22.73 18.26 18.49 17.87 19.83 21.40 19.51 20.77 19.59 19.91 20.22 18.89 21.08 22.18

17.37 18.04 17.62 18.20 18.29 18.54 18.87 17.37 19.20 20.20 20.95 22.69 20.36 20.45 19.87 18.12 18.04 15.71 15.29 14.96 15.04 16.37 17.45

35

Temperature (OC)

30 25 20 Simulated

15

Observed 10 5 0 1

3

5

7

9

11

13

15

17

19

21

23

25

27

29

31

Days

Figure 2: Time series plot for simulated and observed water temperature for January


135


25

Temperature (0C)

20

15 Simulated Observed

10

5

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 Days

Figure 3: Time series plot for simulated and observed water temperature for June

25

Temperature (OC)

20

15 Simulated 10

Observed

5

0 1

3

5

7

9

11

13

15

17

19

21

23

25

27

29

31

Days

Figure 4: Time series plot for simulated and observed water temperature for July From the figures and table, both the simulated and observed water temperature closely follows the same trend and pattern. This close trend observed between the simulated and observed data is further strengthened by the coefficient of determination (R2) value. Coefficient of determination (R2) measures the proportion of the total variation in the data that is explained or accounted for by the regression model (Udom, 2011; Krause et al., 2005; Moriasi et al., 2007). The R2 values for January, June and July are 0.92, 0.83 and 0.6 respectively. Figures 5-7 show fit linear regression graphs using the model predicted values as independent variable and the measured values as dependent variables.


136

Simulated Temperature(oC)


35 y = 1.1509x - 1.8972 R² = 0.9147

30 25 20 15 10 5 0 0

5

10

15

20

Observed Temperature

25

30

(oC)


Figure 5: Regression analysis of simulated and observed temperature for January

25 y = 1.1849x - 3.3899 R² = 0.8289

20 15 10 5 0 0

5

10

15

Observed Temperature

20

25

(oC)


Figure 6: Regression analysis of simulated and observed temperature for June

25

y = 0.6319x + 6.8559 R² = 0.595

20 15 10 5 0 0

5

10

15

20

25

Observed Temperature (oC) Figure 7: Regression analysis of simulated and observed temperature for July


137


4.

CONCLUSIONS

The need for water quality standard, measurement and monitoring cannot be overemphasized. Water temperature is one of the most important physical characteristics. Its importance is seen because of its great influence on the process occurring in the water ecosystem and its effects on other water quality parameters. Hydrologic/water quality models have been shown to be used to monitor and predict water quality parameters. HEC-RAS model was used to simulate water temperature of Owena dam in this study. From the results obtained, it can be concluded that HEC-RAS can actually simulate water temperature of Owena dam in Ondo state of Nigeria. REFERENCES Abbaspour, K. C., J. Yang, I. Maximov, R. Siber, K. Bogner, J. Mieleitner, J. Zobrist, and R. Srinivasan. 2007. Modelling hydrology and water quality in the pre‐alpine/alpine Thur watershed using SWAT. J. Hydrol. Vol. 333(2‐4), Pp. 413‐430. Agoshkov, V. I. 2002. Mathematical Models of Life Support Systems. In V. I. Agoshkov (Ed.), Knowledge for Sustainable Development, An Insight into the Encyclopaedia of Life Support Systems. Vol. 1 UNESCO/EOLSS. Pp 335-281 Ambrose, R.B., Wool, T.A. & Martin, J.L. 1993. The Water Quality Simulation Program, WASP5: model theory, user’s manual, and programmer’s guide. U.S. Environmental Protection Agency, Athens, GA. Amod S. Dhakal, Evan Buckland and Scott Mc Bain 2012. Draft Technical Memorandum: Overview of methods, and results to develop unimpaired, impaired and future flow and temperature estimates along lower Alameda creek for Hydrologic years, 1996-2009” Hydrology and Water Temperature modeling Report, P. 61 Arnold, J. G., Allen, P. M. and Bernhardt, G., 1993. A comprehensive surface-groundwater flow model. Journal of Hydrology Vol. 142, Pp. 47-69. Berit Arheimer and Jonas Olsson 2004. Integration and Coupling of Hydrological Models with Water Quality Models: Applications in Europe” Swedish Meteorological and Hydrological Institute (SMHI) SE-601 76 Norrköping, SWEDEN Bicknell, B.R., J.C. Imhoff, J.L. Kittle, A.S. Donigian, and R.C. Johanson 2001. Hydrologic simulation program-FORTRAN. User‘s manual, v.11, Athens, GA., USEPA. Chapra, S.C., Pelletier, G.J. and Tao, H. 2008. QUAL2K: A Modeling Framework for Simulating River and Stream Water Quality, Version 2.11: Documentation and User’s Manual. Civil and Environmental Engineering Dept., Tufts University, Medford, MA. Gassman, P.W., Reyes, M.R., Green, C.H., Arnold,J.G. 2007. The Soil and Water Assessment Tool: Historical Development, Applications, and Future Research Directions. Transactions of ASABE Vol. 50(4) Pp. 1211-1250. HEC-RAS. 2010. River Analysis System. U.S Army Corps Engineers, User’s Manual for version 4.1. Pp.51-700 Krause P., Doyle, D.P., Base, F. 2005. Comparison of different efﬁciency criteria for hydrological model assessment. Advances in Geosciences, Vol. 5, Pp. 89–97. Love, J. T., and A. S. Donigian Jr. 2002. The Connecticut watershed model: Model development, calibration, and validation. In Proc. WEF Watershed 2002. Alexandria, Va.: Water Environment Federation. Moriasi, D. N., Arnold, J. G., Van Liew, M. W., M. W. Bingner, M. W., Harmel, R. D., Veith, T. L. 2007. Model evaluation guidelines for systematic quantification of accuracy in watershed simulations. Transactions of the ASABE. Vol. 50(3) Pp.885−900 Moriasi, D.N., Wilson, B. N., Douglas-Mankin, K. R., Arnold, J. G. and Gowda, P. H. 2012. Hydrologic and water quality models: use, calibration, and validation. Transactions of the ASABE, vol. 55, no. 4, pp. 1241–1247, Park, R.A. and Clough, J.S. 2012. Modeling Environmental Fate and Ecological Effects in Aquatic Ecosystems. AQUATOX (RELEASE 3.1). Technical documentation, U.S. Environmental Protection Agency, vol. 2 Nigerian Institution of Agricultural Engineers © www.niae.net

138


Pickett P.J. 1997. Pollutant loading capacity for the Black River, Chehalis River system, Washington. Journal of the American Water Resources Association Vol. 33(2), Pp. 465-480. Qinggai Wang, Shibei Li, Peng Jia, Changjun Qi, and Feng Ding 2013. A Review of Surface Water Quality Models. The Scientific World Journal, Vol. 2013 (2013), P. 7 Schwarz, G.E., Hoos, A.B., Alexander, R.B. & Smith, R.A. 2006. The SPARROW Surface WaterQuality Model—Theory, Applications and User Documentation: U.S. Geological Survey, Techniques and Methods 6–B3, P. 248 Stephanie, Mc Caffery. 2008. Water Quality parameters and Indicator. Waterwatch New South Wales, Namoi Catchment Management Authority, Pp. 1-6 Tsakiris G. and D. Alexakis 2012. Water Quality models: An overview. European Water Publications, Vol. 34, Pp. 33-46 Udom, A.U 2011. Simple linear regression and correlation analysis. Essentials of statistics. 2nd Edition. Loius Chumez ptinting Enterprise (Nig). P. 35 Vasudevan, M., Nambi, I.M., and Suresh Kumar, G. 2011. Application of QUAL2K for assessing waste loading scenario in River Yamuna. International journal of advanced technology and Engineering.


139


GUIDE FOR AUTHORS Publication Schedule: The Journal of Agricultural Engineering and Technology (JAET) is published annually (four issues) by the Nigerian Institution of Agricultural Engineers (NIAE), A division of the Nigerian Society of Engineers (NSE). Published articles can be viewed at the Institution’s website: www.niae.net. Manuscript: The manuscript should be typed double spaced on A4 paper (216mm x 279mm) on one side of the paper only, with left, right and top-bottom margins of 25.4mm. The original and three copies are required for initial submission. The paper should not exceed 20 pages including Figures and Tables. Manuscripts are also accepted as attached file in MS word. Organization of the Manuscript: The manuscript should be organized in the following order; Title, Author’s name and address including E-mail address and telephone number; Abstract; Keywords; Introduction; Materials and Methods; Results and Discussion; Conclusion; Notation (if any); Acknowledgements; References. The main headings listed above should be capitalized and left justified. The sub-headings should be in lower case letters and should also be left justified. Sub-sub headings should be in italics. All headings, sub-headings and sub-sub-headings should be in bold font. Headings and sub-headings should be identified with numbers such as 1; 1.1; 1.1.1 etc. For the sub headings, the first letter of every word should be capitalized. Title: The title should be as short as possible, usually not more than 14 words. Use words that can be used for indexing. In the case of multiple authors, the names should be identified with superscripted numbers and the addresses listed according to the numbers, e.g. A. P. Onwualu1 and G. B. Musa2. Abstract: An abstract not exceeding 400 words should be provided. This should give a short outline of the problem, methods, major findings and recommendations. Keywords: There should be keywords that can be used for indexing. A maximum of 5 words is allowed. Introduction: The introduction should provide background information on the problem including recent or current references to work done by previous researchers. It should end with the objectives and contribution of the work. Materials and Methods: This section can vary depending on the nature of the paper. For papers involving experiments, the methods, experimental design and details of the procedure should be given such that another researcher can verify it. Standard procedures however should not be presented. Rather, authors should refer to other sources. This section should also contain description of equipment and statistical analysis where applicable. For a paper that involves theoretical analysis, this is where the theory is presented. Results and Discussion: Results give details of what has been achieved, presented in descriptive, tabular or graphical forms. Discussions on the other hand, describe ways the data, graphs and other illustrations have served to provide answers to questions and describe problem areas as previously discussed under introduction. Conclusion: Conclusion should present the highlights of the solutions obtained. It should be a brief summary stating what the investigation was about, the major result obtained and whether the result were conclusive and recommendations for future work, if any. Notation: A list of symbols and abbreviation should be provided even though each of them should be explained in the place where it is used. References: Follow the name-date system in the text, example: Ajibola (1992) for a single author; Echiegu and Ghaly (1992) for double authors and Musa et al. (1992) for multiple authors. All references Nigerian Institution of Agricultural Engineers © www.niae.net

140


cited must be listed in alphabetical order. Reference to two or more papers published in the same year by the same author or authors should be distinguished by appending alphabets to the year e.g. Ige (1990a, 1992b). All references cited in the text must be listed under section “References”. For Journal, the order of listing should be author’s name, year of publication, title of paper, name of journal, volume number, pages of the article: for books, the author’s name comes first followed by the date, title of book, edition, publisher, town or city of publication and page or pages involved. Examples are as follows: Journal Articles: Ezeike, G. O. I. 1992. How to Reference a Journal. J. Agric Engr. and Technology. 3(1): 210-205 Conference Papers: Echiegu, E. A. and Onwualu A. P. 1992. Fundamentals of Journal Article Referencing. NSAE paper No 92-0089. Nigerian Society of Agricultural Engineers Annual Meeting, University of Abuja, Abuja – Nigeria. Books: Ajibola O. 1992. NSAE: Book of abstracts. NSAE: Publishers. Oba. Abakaliki, Nigeria. Book Chapter: Mohamed S. J., Musa H. and Okonkwo, P. I., Ergonomics of referencing. In: E. I. U. Nwuba (Editor), Ergonomics of Farm Tools. Ebonyi Publishing Company, Oshogbo, Osun State, Nigeria. Tables: Tables should be numbered by Arabic numerals e.g. Table 3 in ascending order as reference is made to them in the text. The same data cannot be shown in both Table and Figure. Use Table format to create tables. The caption should be self explanatory, typed in lower case letters (with the first letter of each word capitalized) and placed above the table. Tables must be referred to in the text, and positioned at their appropriate location, not at the end of the paper. Figures: Illustrations may be in the form of graphs, line drawings, diagrams schematics and photographs. They are numbered in Arabic numerals e.g. Figure 5.m. The title should be placed below the figure. Figures should be adequately labeled. All Figures and photographs should be computer generated or scanned and placed at their appropriate locations, not at the end of the paper. Units: All units in the text, tables and figures must conform to the International System of units (SI) Reviewing: All papers will be peer reviewed by three reviewers to be appointed by the Editors. The editors collate the reviewers’ reports and add their own. The Editor-In-Chief’s decision on any paper is final. Off Prints: A copy of the journal is supplied free of charge to the author(s). Additional reprints can be obtained at current charges. Page Charges: The journal charges a processing fee of N1000 and page charges are currently N1000 per journal page. When a paper is found publishable, the author is advised on the page charges but processing fee (non refundable) must be paid on initial paper submission. These charges are subject to change without notice. Submission of Manuscript: Submission of an article for publication implies that it has not been previously published and is not being considered for publication elsewhere. Four copies of the manuscript and N1000 processing fee should be sent to: The Editor-In-Chief Journal of Agricultural Engineering and Technology (JAET) C/o The Editorial Office National Centre for Agricultural Mechanization (NCAM) P.M.B. 1525, Ilorin, Kwara State Nigeria. Papers can also be submitted electronically to any member of the Editorial Board and to the Technical Assistant to the Editor In Chief at [email protected]. Nigerian Institution of Agricultural Engineers © www.niae.net

141