This is the final post series about PORT PERFORMANCE MEASUREMENT (FOURGEAUD, Patrick; World Bank paper). We hope to be contributed to establish management concepts related this issues and benchmarks.
4- NOTE ON BERTH OCCUPANCY RATE AND QUEUING THEORY:
Ships are berthed according to available space and other constraints as number and size of bollards, number and location of main pieces of handling equipment, nautical constraints etc. For example, a 1000-meter quay can theoretically receive three large Panamax-type ships or four or five smaller ones. Only simulation systems could take these details into account. Analytical approaches require that a number of berths be specified in advance.
4- NOTE ON BERTH OCCUPANCY RATE AND QUEUING THEORY:
Ships are berthed according to available space and other constraints as number and size of bollards, number and location of main pieces of handling equipment, nautical constraints etc. For example, a 1000-meter quay can theoretically receive three large Panamax-type ships or four or five smaller ones. Only simulation systems could take these details into account. Analytical approaches require that a number of berths be specified in advance.
It must be done by considering the length of ships commonly operated. Thus, if a terminal actually accommodates ships with various sizes and berthing space can be optimized, its real capacity may be slightly underestimated when using the methods described below.
4.1- berth occupancy rate
This rate is usually computed over a year, to include seasonal effects:
Cargo handling performance may be monitored by recording:
4.1- berth occupancy rate
This rate is usually computed over a year, to include seasonal effects:
Cargo handling performance may be monitored by recording:
formula:
Cumulated length of commercial operations alongside the quay v/s... /
_________________________________________________...available time over the given period
The optimal use of infrastructure might be best monitored with the following ratio:
formula:
The optimal use of infrastructure might be best monitored with the following ratio:
formula:
Cumulated length of call alongside the quay (including idle time) v/s.. /
________________________________________________
...available time over the given period (365 x 24 h)
...available time over the given period (365 x 24 h)
The difference accounts for all tasks and procedures to be performed when the ship is berthed, at the beginning or the end of the call. It also includes the consequences of the organization of work: restricted working time, lack of flexibility (shifts scheduled at fixed hours) etc., and the consequences of other constraints that apply to ships mooring or sailing out (tide, current, availability of pilots and tugs, swing bridges, locks...).
A distinction must be made according to the way ships are chartered:
Liner-ships have to comply with a precise schedule. If no berth is available at the time of ar
rival, the call may be canceled, the cargo shifted to another port or waiting for the next call. Thus, whenever competition exists between ports, the berth occupancy rate usually does not exceed 50 to 60 %.
Higher rates can be seen when port facilities are saturated and there is no alternative, or when it is possible to schedule precisely a high number of calls; e.g., terminals dedicated to a single intensive traffic like short-sea Ro/Ro traffic or some private terminals operating on East-West trade with very intensive and well coordinated activity.
Chartered ships are usually less affected by port congestion, depending on the nature of the cargo, the demurrage rate. Calls may be planned only a few days or weeks in advance.
Their length may vary according to the nature and the size of the shipment. High berth occupancy rates can be observed, up to 80%, sometimes more, generating significant waiting time.
Expanding the working period to 3 shifts per day and to the week-end, whenever possible, is the first and simplest way of improving this ratio.
4.2- queuing theories and simulations
Port congestion can be precisely assessed by using simulation models taking into account each significant step of the process. In some case, rough estimates can be obtained through simplifying methods. A common one is based on the computation of the ratio:
formula: waiting time/operating time, according to the berth occupancy rate and the number of berths[1].
Theoretical requirements are: a set of identical berths where an homogenous fleet of ships call on a first come first save basis. Arrival patterns and distributions of service times are approximated by a statistical law (Erlang distribution), simulating processes ranging from the random distribution (Erlang 1) to increasingly regular ones (Erlang 2, 3...).
· The average assumptions for a freighted traffic are:
no distinctive pattern of calls => arrival at random,
- and various types of cargoes => service time at random.
- or homogeneous traffic => more regular service time, increasing order of the Erlang law,
· The average assumptions for liner traffic are:
- strict compliance with a schedule, i.e., a fixed distribution pattern of arrivals, => (Erlang ∞), practically Erlang law at 2 to 4th order.
- variation of service time depending on the nature and size of shipment. => increasing order.
This method is more suitable for chartered traffic and should not be applied to liner ships, as long as they do not wait. In such a case , the max berth occupancy rate depends primarily on the actual possibility to schedule calls evenly distributed.
This method represents correctly two phenomena:
- and various types of cargoes => service time at random.
- or homogeneous traffic => more regular service time, increasing order of the Erlang law,
· The average assumptions for liner traffic are:
- strict compliance with a schedule, i.e., a fixed distribution pattern of arrivals, => (Erlang ∞), practically Erlang law at 2 to 4th order.
- variation of service time depending on the nature and size of shipment. => increasing order.
This method is more suitable for chartered traffic and should not be applied to liner ships, as long as they do not wait. In such a case , the max berth occupancy rate depends primarily on the actual possibility to schedule calls evenly distributed.
This method represents correctly two phenomena:
- a very rapidly increasing waiting time when the berth occupancy rate rises;
- a very rapidly decreasing waiting time when the number of berths increases; dedicating an additional berth to an existing traffic improves flexibility and capacity to a much larger extent than the mere relative increase in the number of berths.
Generally, the maximum berth occupancy rate (corresponding to a low average waiting time) is lower with liners than with freighters.
Synthesis of the proposed approaches:
FINAL
[1] These methods, initially much developed for telecommunication purposes, have been extensively applied to port planning and design in the seventies. cf. UNCTAD Manual on Port Management 1976; the number of interchangeable berths where the same ship can call.
