Future Storage Area Requirements for Sustainability of Izmir Container Port

3 (2), 2009, 161 - 164

©BEYKENT UNIVERSITY

Future Storage Area Requirements for

Sustainability of Izmir Container Port

Doğan CANIVAR, Ümit GÖKKUŞ and Adem EREN

Celal Bayar University, Manisa, Turkey adem.eren@bayar.edu.tr umit.gokkus@bayar.edu.tr Received: 16.02.2007, Accepted: 18.11.2009

Abstract

Inventory method is well known in management and econometrics. But in port planning, it is rarely used because of yielding roughly the solution resulting from assuming average holding days of containers at yard and one-day data representing whole days in a month. In this study, the inventory model is utilized for future storage requirements of Izmir port. Especially in city ports, the size of storage area needed affects directly the city traffic and planning. This study gives the general information whether the size of storage area required for future extension of port is adequate for the allowed development area in land side of port. The exact solution can be found by using complicated mathematical modeling and simulation.

In this study, especially the inventory and cost model are processed for present values of container statistics. The similar process is maintained in order to determine the future storage area of Izmir port. Even though the limited data is analyzed, the reasonable solutions are obtained by assuming some parameters such as storage days and forecasting results. For three- five-year planning terms, the inventory and cost model is utilized for future extension of storage area existing in Izmir port.

Keywords : Container storage, storage yard optimization, port planning, container traffic, inventory method

ÖZET

Izmir Konteyner Limaninin Sürdürülebilirliği Için Gerekli Stok Sahasi Ihtiyaci Envanter yöntemi, işletme ve ekonomide iyi bilinen bir yöntemdir. Fakat, bir aydaki bütün günler için temsilen bir günü alması ve stok sahasında konteyner tutma sürelerini ortalama bir değerle temsil ettirmesi gibi kaba yaklaşımı ile liman planlamasında çok sık kullanılmamaktadır. Bu çalışmada envanter yöntemi İzmir Limanının gelecekteki stok sahası ihtiyacı hesabı için

trafiğini ve planlamayı etkilemektedir. Genişleme için gerekli stok sahası boyutunun, gelişme için izin verilen liman kara sahası alanı için uygun olup olmadığı konusunda bu çalışma genel bilgi vermektedir. Tam çözüm, karmaşık matematik modelleme ve simülasyon ile bulunabilir. Çalışmada, özellikle envanter ve maliyet modeli bugünkü konteyner

istatistiklerini için çalıştırılmıştır. Benzer işlem İzmir Limanının gelecek stok sahası ihtiyacının tespitinde de yapılmıştır. Verinin sınırlı olmasına rağmen, tahmin sonuçları ve stokta bekleme süreleri üzerine bazı kabuller yaparak kayda değer sonuçlar alınmıştır. İzmir limanının mevcut stok sahasının gelecekteki genişleme miktarı için, limanın üç-beş yıllık planlama dönemleri boyunca envanter ve maliyet modeli kullanılmıştır.

Anahtar Kelimeler: Konteyner stok sahası, Stok sahası optimizasyonu, Liman planlama, Konteyner trafiği, Envanter yöntemi

Introduction

Izmir city port is one of the major container ports in Turkey, where is located at Izmir Bay in Aegean Sea. The port area is limited since it is surrounded by the city. However, the demand for sufficient berth length and container storage area is essential for a port to conduct its functions. In Izmir port, the most important problem is to determine both the optimum number of berth lengths and container storage area required for handling and storage capacities in future. The present handling capacity of port can be determined according to the existing cargo and ship statistics. For future expansion of the port, the demand for port cargo and ship traffic must be forecasted by statistical analyses on the time series representing the seasonal variation of incoming and outgoing cargoes, the volume of domestic and international trade and number of ships arriving at the port.

There are two important factors forcing port management to carry out the required port activities in the limited area considered for future extension. They can be expressed as increasing the traffic congestion between marshalling yard of port and urban traffic and exceeding the target values estimated for the cargo volume and number of ships arriving at the port. Even though Port of Izmir has the limited extension area required for future demand of container storage, all containers coming and outgoing to the port must be handled. This area are limited and only extended up to 320 000 m2 (present area to be filled possibly). The problems arising from the limited storage area and the increasing cargo should be reconciled by determining the optimal storage duration and selecting conveniently the container handling techniques in storage area. The storage area can be computed by several approaches such as the conventional techniques [11], UNCTAD (United Nations Conference on Trade and Development) graphics especially to be used for ports of developing countries and inventory models including the

deterministic and probabilistic fundamentals. In this study, the size of storage area for container is calculated by considering the deterministic inventory theory together with selecting the straddle carrier system for container handling [1]. To carry out this process, the required data is composed of the statistics regarding the number of containers recorded at the storage area in a certain day of each month in a year and the holding duration of each container at the storage area. Other parameters to be considered are the required storage area per container, number of container-holding day, annual amortization rate and income per container for handling and holding.

The aim of this study is to determine the optimum container storage area using the deterministic inventory model and port economy covering the refund of investment cost per idle container slots and deprivation of storage cost, loss of storage cost due to unemployed container slots or queuing cost, for ships at the berth, railcar at marshalling yard and trucks at inland depot having to be waited at queue because of overall capacity of storage yard. Cost analysis of this study is mainly based on the refund of investment cost and deprivation of storage cost. By minimizing these two costs, the optimum container capacity can be determined together with the optimum storage area including the certain stacking height of container. This study gives the optimal solution for future extension of container area with linear transformation of the monthly-stored values of the container forecasts from the macro and micro projections. Finally, the optimum storage size and the number of day recorded as full storage capacity can be determined by considering some assumptions in forecasting and modeling.

Inventory Method

The inventory method is one of the mathematical optimization methods. This aims to compute the number of idle slots and containers waiting for access to the storage yard with full capacity. By adding the cost analysis to the

mathematical model, the economical and physical optimization of yard can be performed. Therefore, container terminals can be planned in accordance with the productivity of handling equipment, the optimum length of quays, and the minimization of idle container slots.

The principle of inventory method to be used as optimization method is to develop a stochastic stock planning model fitted to arrival and service distributions of containers. This model can be established as deterministic and stochastic process. The deterministic models suppose that the data on demand, capacity and unit costs (e.g. cost of invested area per square meters, cost of waiting container per day) are known. But this model can be only managed by gathering the sufficient number of statistics. On the other hand, uncertain data can be also evaluated by stochastic inventory models based on the probability distributions identified by processing the known data. For this reason, the

stochastic inventory model can be developed by using the demand data represented by the tested probability distributions, capacity and unit costs.

The seasonal variation of container traffic affect directly the port economy, the size of storage area, the required handling equipment, berth capacity, port management and operations. A low efficiency in storage area may occur either with increasing containers waiting for service due to the container congestion born in insufficient stocking area or existing idle container slots due to the excessive capacity in container storage area. The containers exceeding port storage capacity have to be waited at the berth, marshalling yard and inland depot or another storage area within port until the container slots will be emptied. The optimization methods give the possibility to yield optimally the required area for storage whereas the refund of investment cost together with the queuing costs for containers waiting at the queue for storage or the deprivation of storage costs due to transit containers leaving port to take better service should be minimize.

J1 Cost

o System Capacity Optimum Capacity

Figure.1 Container terminal stock yard capacity-cost

function

It is seen in Figure. 1 that the curve on deprivation of storage cost become higher and steeper with increasing load demands at stock yard whereas the curve on refund of investment cost is reduced. On the other hand, the curve of total cost becomes lower to optimum capacity point and higher after this point. The minimum total cost corresponds to the optimum capacity of storage yard.

INVENTORY MODEL FORMULATION FOR IZMIR PORT

In this inventory model, there are two most important factors: number of containers staying at stock yard together with their holding durations and storage costs of containers at yard. They can be expressed as follows.

F=f0.QS.t0 (m2) (1) Css

=—

C.F

365

Css = Cf .C.fQt 0 (3)

365

where F; stock yard area required for holding durations (to), Qs; daily container capacity of yard (T E U/ d a y ) {here TEU is a container box size called for twenty feet equivalent unit, forty feet containers are called 2 TEU}, f0; unit area per container box (m2/TEU) for second (15 m2), third (10 m2) and fourth (7.5 m2) stacking height, C s s ; daily terminal stocking cost (U S D/ d a y ) , depending on terminal substructure, investment capital, interest rate and amortization factor, C; unit cost of yard (USD/m2), Cf ; annual amortization rate {(0.111 U S D/ d ay) , annual interest rate (r=%2), life of investment (N=10 years) and the unit cost of idle slots, C s =(0.300 U S D/ T E U . d a y ) }.

Objective function of the total cost which contains the refund of investment costs of idle slots as first term in case of Qs>Qi and the deprivation of storage cost as second term in case of Qi >Qs can be expressed as [11]:

S n

C(Q,

) =

2

- Q ) t + C

2

- Qs)

i=1 i=S+1

where Cb is unit deprivation of storage due to loading and unloading, terminal, warehousing and assisting trailer services (39.64 TEU/day). To use a holding duration recorded for each container may cause more complex situations , so an average holding day to is taken as a holding duration ti. This means that the refund of investment cost will happen for each idle container during an average holding day to in case of Qs>Qi .

Case Study: Inventory Model Application On Izmir Port

According to the statistics on Izmir port, the number of containers countered in the certain time of each month at yard is seen in Table. 1. Relation between handling capacity and its holding duration is represented as in Figure.2a

belonging to Haydarpasa port which is similar to Izmir Port from port

organization and administration point of view. Both of ports are administrated by same port authority (TCDD governed by State). This figure is drawn for handling capacity, 250 000 TEU, 300 000 TEU and 350 000 TEU according to

1998 Statistics, TCDD.

Table.1 Recorded container number (TEU) at yard [12] Months Total No TEU

Jan. 5140 8620 Feb. 5670 9253 Mar. 6320 10140 April 6885 10814 May 7540 11725 June 8654 13245 July 9947 15042 Aug. 11076 16614 Sep. 11816 17156 Oct. 12241 17945 Nov. 12864 19227 Dec. 13004 20145 UI H •J _«I n

Hil-i^Piv l--.---.-iMi>'..

-al-ia) (b) Figure.2 Holding duration versus handling capacity

By using the numerical analysis, the handling capacity-storage days relation for handling capacity of 480 000 TEU and 700 000 TEU are developed by considering the 5-day and 15-day holding duration as extreme points of graphics (Figure.2b). For 5-day, For 15-day,

Vi

y

where y1, y2; storage capacities for the extreme points corresponding 5-day and 15-day holding duration, x; annual handling capacity. According to handling capacity x forecasted for future demand, the storage capacities for the 5-day and 15-day extreme points are firstly calculated and then the constants a and b are linearly stated as seen in Table.2 to obtain the storage capacity

corresponding to any holding days.

Table.2 Linear form between holding duration and storage capacity, y =a x+b Years 2002 2005 2010 2015

a 1532.7 1529 1667 1528 b 2846.5 2846 2868 6490

where y; storage capacity and x; storage or holding duration are taken into consideration. In 2002 years [12], the handling capacity is stated as approximately 507 000 TEU. Storage days, to, considered for each various container capacity, Qs, are calculated by employing the mentioned relations. The proposed model to Izmir port is applied as presented in the following tables.

Table.3 Number of full and vacant container slots of the 14 000-TEU capacity and 7 day holding duration in 2002

Qi Qs-Qi Qs-Qi Qi-Qs

8620 5380 5380 0 9253 4747 10127 0 10140 3860 13987 0 10814 3186 17173 0 11725 2275 19448 0 13245 755 20203 0 15042 -1042 0 1042 16614 -2614 0 3656 17156 -3156 0 6812 17954 -3954 0 10766 17945 -3945 0 14711 19227 -5227 0 19938

Table.4 Various container capacities versus their total costs No Qs-Qi Qi-Qs Cdeprivation Crefund CTotal Qs t0

1 20203 19938 832768.6 554960 1387729 14000 7 2 26203 13938 615389.5 594600 1387729 15000 8 3 33161 8896 442172.1 634240 1076412 16000 9 4 40547 4282 291379.5 673880 965259.5 17000 10 5 49492 1227 211961.9 713520 925481.9 18000 11 6 60492 227 244917.1 753160 998077.1 19000 13 7 78265 2 328713 812620 1141333 20500 14 The total cost values versus yard capacities in Table.3 and Table.4 are graphically presented in Figure.3.

Figure.3 Total cost versus yard capacity

By varying storage capacity, Qs, the total cost depending on monthly-recorded container numbers, Qi, and stocking days, t0 can be computed by utilizing the objective function. The stocking yard capacity corresponding to the minimum total cost is therefore stated as optimum.

When the following procedure is maintained, macro and micro projection results to the future extension in Table.5 provide the required inputs for calculating the average storage days of each considered storage capacity and then the optimum storage capacity of yard and their minimum total costs are presented as in Table.6. As seen in this table, additional storage capacity to be allowed for the four level stacking causes about 15% growth for first five-year planning term, and about 45% growth for second term according to the results of inventory method.

Conclusion

This study aims to determine the optimum size of container storage area in Izmir Port especially with using inventory model. For this purpose, the limited data are supplied from the relevant port authority. This data should be

applicable to request of the inventory method. It is seen that the inventory model yields reasonable outputs for determining size of storage area at yard. The most important parameters such as the holding duration of containers, number of recorded containers staying at yard and unit investment cost per slot, unit service cost per container and amortization parameters should be adequately determined before processing the optimization model. The seasonal variation of container traffic and the daily movement of containers staying at yard should not be considered in this study.

The optimum values must be among data series of the container numbers at yard. This affects usage frequently of this model.

Table.5 Monthly container number countered at yard for future extension, according to forecasts obtained from macro and micro projections

Months 2005 2010 2015 January 8 074 11 385 16 359 February 8 515 12 007 17 251 March 11 157 15 733 22 605 April 11 451 16 147 23 200 May 11 451 16 147 23 200 June 11 011 15 526 22 307 July 10 864 15 319 22 010 August 11 305 15 940 22 902 September 16 002 22 564 34 420 October 15 855 22 358 32 122 November 14 681 20 702 29 743 December 15 855 22 358 32 123

Table.6 Optimum storage area for future extension of Izmir Port Years Handling

Capacity y1 Y2 t0 Qs,Opt. CTotal,Min 2005 695 184 10 493 25 787 8 19 000 744 193 2010 980 785 11 201 27 867 11.5 22 000 1 127 039 2015 1 408 147 14 128 29 404 16 32 000 1 798 511 This inventory and cost model gives the general idea about size of container storage area. The data on Izmir port is analyzed and the reasonable outputs are yielded. The existing capacity is 18 000 TEU in 2002. Future storage

capacities in 2005, 2010 and 2015 reach respectively 19000, 22000 and 32 000 TEU.

REFERENCES

[1] ALKINS, W. H., (1995) Modern Marine Terminal Operations and Management, Ed. R.A. Boyle, Port of Oakland

[2] EREN, A., GOKKUS, U., (1997) Container Traffic Forecast and Storage Capacity of Izmir Port, Advances in Civil Engineering, Third Technical Congress, Proceedings,

Hydraulic Engineering Vol. III, Middle East Technical University, p.731-740

[3] GOKKUS, U., EREN, A., (1998) Determination of Storage Area Size and Configuration of North-Aegean Container Port to Container Traffic Capacity, Third National Conference on Turkish Coasts and Near-shore, Middle East Technical University, Proceedings, Ankara, p.805-816

[4] HAMDY, A. T., (1997) Operations Research: An Introduction, Prentice Hall [5] HILLIER, F. S., LIBERMAN, G. V., (1974) Operations Research, Holdenday Inc., San Francisco

[6] JANSSON, J.O. & SHNEERSON, D., (1982) Port Economics, MIT Press

[7] KOZAN, E., (1997) Comparison Of Analytical And Simulation Planning Models Of Seaport Container Terminals., Transportation Planning And Technology, London, v.20, p.235-248

[8] OKADA, H., (1990) Port Planning and Development, the Overseas Coastal Area Development Institute of Japan (OCDI), Text Book, Japan

[9] OZEN, S. & OZMEN, I.H., (2000) A Study on Determining of the Optimum Container Yard Capacity, Third National Coastal Engineering Symposium, Proceeding

Book, 309-324, Canakkale

[10] DEGENAIS, G. M., MARTIN, F., (1985) Forecasting Containerized Traffic for the Port of Montreal (1981-1995), Research Project, University of Montreal

[11] United Nations Conference on Trade and Development (UNCTAD), (1978) Port Development: Handbook for Planners in Developing Countries, New York

[12] General Directorate of State Railways, Turkish Republic (TCDD), (2002) Izmir Regional Directorate, Statistics on Ships, Cargo and Storage Area Operations in Izmir Port (1990-2002 years), Izmir

[13] EREN, A., Optimization of Container Yards, Dokuz Eylul University, Institute of Natural and Applied Sciences, Ph. D. Thesis, Izmir, 2003