The ocean carriers’ key phrase these days seems to be the cliché ‘bigger is better.’ The economics of carrying more containers on fewer ships regardless of the infrastructure consequences has kept ports on a seemingly never-ending path of ‘catch up.’ Port authorities and terminal operators are caught in the middle of a vice the jaws of which are the growing size of ships due to alliances and mergers on one side and on the other side the need to keep container volumes increasing to pay for the requisite infrastructure changes. One thing is obvious in the ocean transport environment: ship size will continue to drive the size of the cranes necessary to serve them. Figure 1 shows the evolution of container ships since their debut in the 1960s.
There is even talk of a 15,000 TEU vessel sometime in the future. Is there no end to how large a ship could be designed? Aside from physical harbour constraints, another impediment that may slow down the size increases of these giants could be propulsion power. Recently-built 6,600 TEU vessels require 90,000BHP (67,000kW) to maintain a transit speed of 25 knots. Vessels beyond 8,000 TEU will need more than 100,000BHP to accomplish this. Some believe that engine manufacturers will find it difficult to squeeze this power from present propulsion designs. Even so, there is little doubt that vessels well over 8,000 TEU will eventually materialise.
Shipping lines are seemingly in a position to ‘have their cake and eat it too.’ Despite the larger dimensions, almost doubled compared to 40 years ago, they’re demanding fast turnaround times for their vessels since time is money. From a terminal operator’s perspective, there are only two ways of improving the turnaround time for a container vessel: employ more cranes or reduce the average cycle time of the cranes. For most terminal operators the second approach would seem more attractive, requiring less capital. Yet the overall crane dimensions and travel distances from ship-to-shore seem to make reduction of cycle time an uphill battle. Witness the crane size increase over the same period since the advent of containerisation.
With 7,900 TEU ships on the water and 8,000+ TEU vessels on the drawing boards, terminal operators have opted for an even larger crane, a fifth generation behemoth known as the Suez class. The Virginia Port Authority recently purchased three of these cranes from ZPMC illustrated in Figure 3. Each has an outreach of 201ft (61.3m) capable of working a 22 container-wide vessel, with boxes stacked six high on deck. To accommodate a higher stack on deck, the maximum clearance from spreader to dock was designed at 120ft (36.6m).
In attempting to maintain productivity, crane purchasers have demanded huge increases in accelerations and speeds to accompany the large dimensional increases and heavier loads such as when handling twin 20ft containers. Typical crane machinery dynamics are shown in Table 1.
No pain, no gain
Escalating ship sizes has resulted in a variety of costs to the terminal operator. The costs to replace wires and sophisticated electronic components have risen markedly. Accessing remote parts of the crane for maintenance is a challenge. Wharves have to be stronger, with more concrete and steel, and with higher construction costs. Spring effects of larger cranes have resulted in phenomena such as harmonic oscillations to be considered in the design, usually resulting in more steel and adding to the weight of the crane.
Beyond these considerations are the detrimental effects on the operation of the crane. Longer hang lengths result in greater rope stretch, catenary effects, and longer pendulum periods, which take longer to damp. The operator is now twice as far from his work as he might have been 40 years ago, increasing parallax and depth perception effects. And he must move containers further from the ship to dock and vice versa. Can he maintain or improve productivity despite the problems associated with the larger, faster cranes? There has been speculation in the industry that cycle times will grow and individual crane productivity will diminish.
The Port of Virginia is in an ideal position to study such a problem since the crane inventory consists of machines from all generations, from the first old Paceco A-frame delivered in 1969 to the largest cranes yet delivered, the ZPMC-built Suez class. In this study I incorporated two tools to analyse crane capability and performance, a crane simulation programme developed by Liftech Consultants of Oakland, California, and a crane monitoring programme developed under the guidance of Tony Simkus, Virginia International Terminals’ (VIT) assistant director of engineering and maintenance.
Computer simulation
Simulation has been used quite effectively in evaluating and designing many aspects of the terminal operation from the wharf to the yard to the gate. Using this tool on the ship-to-shore operation, one can vary certain physical parameters such as crane and ship dimensions, spotting times (referred to as dwells), crane accelerations and speeds, bay cross sections and container locations, location of working lanes on the wharf and the horizontal and vertical location of the ship relative to the wharf. Liftech’s programme is also capable of adding delays such as hatch covers and inter-box connectors (IBC’s), as well as incorporating various specialty crane configurations such as dual hoist and the elevating girder. For this study I incorporated two typical bay cross sections, a 4,500 TEU Panamax vessel and a Suezmax ship viewed on the computer screen as shown in Figures 4 and 5 on the next page. The Panamax size was chosen since it comprises the majority of larger vessels serviced at VIT. The Suezmax size was selected to analyse the productivity effects when using the fifth generation crane once those ships are placed into service.
In the simulation, actual measured dwell times along with measured times for hatch top covers and IBC placement and removal times were incorporated. Other delays such as mechanical downtime, waiting for equipment arrival, changing bays, or changing operators, while very important for productivity purposes, were not included in this study to simplify the comparison of computer data with actual measured crane cycle times. I used the simulated productivity using cycle times, incorporating only hatch cover and IBC delays, as the theoretical maximum ship-to-shore productivity against which to compare data taken by monitors.
Production monitoring
To validate the computer simulation, individuals trained as crane monitors collected actual data. In order to reduce the tedious process of using pencils and stopwatches as well as minimising errors, handheld computers were used. More than 15,000 records were collected encompassing second, third and fifth generation cranes in VIT’s inventory. The only crane operations recorded were for those employing a straddle carrier transport operation. For the purposes of this study and to compare the theoretical handling rate (or cycle time) determined through simulation against that recorded empirically, the actual recorded data was purified to include only single load or discharge cycles (which included the empty spreader portion of the cycle) with any associated IBC and hatch cover delays.
Crane size effects when working Panamax vessels
To assess the theory of productivity degradation with increasing crane size, I first used simulation of a third generation versus that of a fifth generation crane working a full discharge and load of the largest bay of a Panamax ship.
The first simulation used the same spotting dwells for both cranes with the appropriate speeds, accelerations, and crane dimensions listed in the table below. It’s obvious that if the operator were at no disadvantage by being positioned in the larger crane that the theoretical productivity would far exceed that of the older, smaller crane merely because the crane was faster. In reality, however, the operator is subject to some degradation because of the longer spreader hang length which results in a longer swing period and also because of visual parallax effects. I attempted to quantify what these effects would be to revise the simulation. Knowing that the pendulum period of a swinging load is proportional to the square root of the hang length, the difference in swing period can be calculated.
If one observes an operator in production, a typical damp time is about one period. This is not hard and fast since some of the tricks of the trade involve minor intentional collisions with fixed objects to stop the sway. One might qualify this assessment by claiming that anti-sway systems can eliminate any operator damping, yet even the best anti-sway systems are not perfect at eliminating load swing. Additionally, operators at VIT and many other terminals prefer not to use installed electronic anti-way. For purposes of this study I estimated that sway damp and parallax effects would amount to about one period damping time plus an arbitrary one second for parallax, an estimated total of 2.5 seconds on each end of the cycle. The dwell parameters were thus reduced for the third generation crane, the simulation rerun, and the following results obtained: Assuming that my estimates were in the ballpark and the simulation sufficiently accurate for the parameters used, the analysis demonstrates that the dynamics of the more modern crane overcome any disadvantages that a dimensionally larger crane would impose on the operator.
Evaluation of actual data
The next challenge was to record and evaluate real operational data under varying conditions to see if the simulation results could be supported. The simulation used a homogeneous bay plan, something rarely found in common practice at multi-user terminals such as VIT operates. Also, a variety of delays to the operation are encountered and operator proficiency varies. However, if a sufficiently large sample of data is taken and this data could be broken down to exclude certain delays, a reasonable comparison of the simulated versus actual measured cycle time could be made.
Over the course of nine months this data was recorded on a Newton handheld touch screen computer, then downloaded into a Microsoft Excel file for further analysis. Only the delays for IBCs and hatch covers, along with the single container loads and discharges, were considered, to compare this data closely with the simulation analysis. A histogram of cycle time frequencies is presented on p26. The actual data shows a difference in the mode or most common cycle time equal to the equivalent of 4.0 containers per hour. This correlates very nicely with a difference of 3.2 containers per hour for the computer simulation for the Panamax vessels, supporting the results which indicate that, at least for the fifth generation crane, we have not reached the point of diminishing returns for operator productivity.
Operator perceptions
Conclusions can be reached by taking data and doing simulations, but feedback from operators who spend many waking hours at the crane controls is critical to the equation. I interviewed several of VIT’s most proficient operators querying them on the differences they have seen in their own productivity as the evolution of crane size and speed has progressed. The opinion was unanimous that although the cab was farther from the spreader, and the wires longer from trolley to spreader, the faster speeds and system responsiveness made up for any parallax or sway period differences. Responsiveness of controls seemed to be a major factor in their favourable opinion. From an engineer’s perspective, this faster response is a combination of increased accelerations and the ability of modern solid state PLC’s to scan the programme permissives faster, resulting in quicker drive commands to the motors.
The operators felt that they would not be concerned about further size and speed increases beyond the fifth generation machine. They also said that anti-sway systems designed to reduce the damping time in the spotting manoeuvres were only marginally effective at reducing cycle time and that there was no substitute for experience and the ‘feel’ of the crane.
What to expect for 11,000 TEU vessels
Since we have validated the simulation data we can now use the simulation programme with the fifth generation crane working the vessel cross section shown in Figure 5, using the same dimensional data and dwells as used for the Suez class crane in the Panamax example above. The handling rate results shown top right indicate a reduction of almost three containers per hour as we might expect due to longer travel paths of the spreader.
Assuming that further increasing accelerations and speeds on this crane would not materially affect operator productivity, for future crane purchases we probably would want to increase the dynamics of the machine to have at least as high a handling rate as that for a Panamax vessel. But then which of the accelerations and speeds should be adjusted? Hoist? Trolley? Both? By how much? Again, simulation comes in handy. By changing the dynamic parameters in the simulation program by constant percentages we can see the relative effects of each.
It is obvious from this comparison, shown below right, that hoist increases give us ‘more bang for the buck.’ Based on this analysis and adjusting the parameters incrementally, I arrived at a set of parameters that would yield the same cycle time or handling rate on the 11,000 TEU bay as that experienced over the Panamax vessel. Since the trolley acceleration affects the handling rate the least, yet has the most effect on operator comfort, I made no change in its value. A comparison is shown in Table 5 (right).
I would recommend for our future procurements that motor powers be increased accordingly.
Conclusion
The combination of computer simulation, operator experience and intuition, and analysis of actual data, shows conclusively that, at least for the size of crane which is needed to service vessels up to and including 11,000 TEU, that boosting crane power along with the associated lift height and outreach should not of itself reduce ship-to-shore cycle time. One must keep in mind that the numbers I have used for this paper may not necessarily resemble those for other terminals. Such crane variations as machinery-on-trolley as opposed to the ones analysed here that use a rope tow crane design can make a significant difference in power requirements. The conclusions reached also assume a standard A-frame design with a boom cannot be raised and lowered. An elevating trolley girder design has many advantages over the fixed design. Computer simulation supports this theory. The practical application is now undergoing research by ZPMC. One thing is for sure – we will be continually challenged by ocean transport lines and the competition with other ports to keep their business. The optimum design of ship-to-shore cranes involves constant research and analysis with the tools available to stay competitive.
