Enhanced weather technology saves fuel
Two companies have recently enhanced their voyage optimisation software offerings in order to focus on fuel savings rather than just safety- and weather-based considerations.
Applied Weather Technology (AWT) of the US has made substantial enhancements to its Bon Voyage system (BVS), intended to help find safer and more fuel-efficient routes. The latest version, BVS 7.0, includes, in addition to access to high-resolution weather and ocean data via broadband, enhanced voyage optimisation and new safety features, and an improved optimisation algorithm that includes customisable speed down and consumption curves that are said to deliver more accurate estimates of fuel cost and time en route.
BVS generates compressed colour-enhanced maps and graphics allowing captains to easily view and interpret potential problem areas. With this data, captains have access to 16-day forecasts updated four times per day and other parameters such as pirate attack information, port vicinity forecasts, satellite imagery and high seas bulletins delivered in near real time. BVS includes several safety features such as resonance alerts to inform captains of the potential for severe motions, easy access to rogue wave forecasts and advanced environmental modelling that enables AWT to forecast storm strengths and effects.
Dr Henry Chen, chief naval architect at Jeppesen Marine, believes that the time has come to replace outdated weather routing with full-scale voyage optimisation software. He believes that weather routing has reached its limitations as shipping companies attempt to cut fuel consumption by slow steaming and virtual arrival.
He says that regulatory requirements coupled with higher fuel prices and low freight rates have now placed the spotlight on fuel efficiency, ahead even of safety. Post voyage analysis has compared the historical passages of ships that sailed using weather routing to see how they would have performed, in terms of saving fuel if they had used voyage optimisation technology instead. The results show a potential fuel saving of up to 5%. So ship owners can not only yield a higher return on their investment through increased operating efficiency and safety, but also reduce GHG emissions. Unlike traditional weather routing with its inherent limitations, voyage optimisation supports current operational tactics such as slow steaming and virtual arrival.
Dr Chen points out that a ship slows down either involuntarily due to increased resistance from the wind and waves or voluntarily due to navigation hazards or fear of heavy weather damage from excessive motion, propeller racing, slamming or boarding seas. Some weather routing systems are now trying to develop ship response prediction capabilities.
To predict ship motion, tools digitise the ship’s body plan, bilge keels and other appendages as well as calculate its loading conditions in terms of fore and aft draughts and GM. A hydrodynamic program then computes added mass and damping coefficients and solves the equation of motion. The results are the so-called Response Amplitude Operators (RAO), which are combined with forecast wave spectra to predict ship responses. Only during the last few years has standard onboard computer hardware had the power to run such computationally intensive programs quickly enough to provide meaningful seakeeping guidance and safe operating limits.
Engine overload is another important parameter missed by basic weather routing systems, says Dr Chen. Shipyards use sophisticated finite element models and high tensile steels to reduce weight and costs in order to be competitive. Similarly, propulsion systems are often optimised for calm weather trial conditions in order to satisfy IMO EEDI requirements. One such design consequence for low speed diesel engines with direct-drive fixed-pitch propellers is high pitch coupled with minimum acceptable sea margin. With calm weather and a clean hull, lightly loaded, the vessel easily makes the contracted speed and good EEDI. Unfortunately, such a practice can lead to frequent engine overload when the ship encounters high wind or seas or increased hull fouling as ships age. If such events cannot be predicted by the weather routing tools, this can lead to over-optimistic ship speed and mistaken diversion decisions when facing heavy weather, not to mention inaccurate estimates of fuel consumption and time of arrival.
There has been substantial research over the years in the area of ship routing algorithms. Most of the weather routing software uses variations of the Dijkstra’s Algorithm, in which the program simulates a vessel departing with full power toward the arrival port with different headings. After each time interval (e.g. six hours), the ship’s dead-reckoned position forms a so-called “isochrone” until it arrives at the destination. A route is then traced back from the earliest arrival time, and fuel consumption is estimated. The claim is that minimum time results in minimum fuel consumption. The calculation is fast, especially when only using speed reduction curves and not taking into account ship motion responses or engine overload.
Unfortunately, the problem with such an approach is the algorithm ignores one important option: speed management. As storms move across the ocean, it is possible for the ship to slow down and let them pass and then catch up, instead of sailing a longer distance to go around, or ‘hove-to’ in bad weather. Such a strategy not only significantly affects fuel consumption for a given arrival time, it reduces the risk of heavy weather damage when fully implemented with ship response and engine overload. The simple speed reduction curves used in weather routing algorithms cannot find the optimum route since they operate only in one dimension. Without modelling the ship performance from first principles, it is not possible to minimise the fuel consumption for a given arrival time without exceeding the safe operating limits.
Speed and heading are both integrated into route optimisation, making the problem multi-dimensional. An algorithm called 3-D Dynamic Programming can be used to minimise fuel consumption for a range of arrival times subject to the constraints of safe operating limits imposed by the captain. The optimisation is performed on a user-defined grid for safe navigation. This allows trade-off between fuel costs and arrival time in a complex inter-modal transportation network. The computation effort is obviously greater since the algorithm must evaluate thousands of speed and heading options compared simpler weather routing. With computer processor speeds doubling every year, the time to solve the problem is reduced from several hours to a few minutes, including the full implementation of ship responses and engine overload. Such systems can be implemented at routing centres shore-side, or onboard ships with daily updates of forecast environmental conditions via satellite communication.
According to Dr Chen, another poor strategy often used by weather routing services is advising the ship to use minimum rpm to arrive on time, since horsepower increases with the cubic power of ship speed. However, this only makes sense in calm weather, because in severe head sea and wind conditions, the added resistance from the wind and waves could result in 50% or more horsepower than that required for calm sea in order to maintain speed. With voyage optimisation, the speed profile on a single route is optimised to arrive on time by slowing down in head seas and speeding up in beam/following seas to catch up after a storm passes. In addition, using a minimum rpm or engine power strategy can preclude better options, such as passing in front of a storm.
Weather routing typically relies on only one weather forecast source, adds Dr Chen. Modern technology has significantly improved forecast accuracy over the past decade, allowing national centres to predict wind and wave forecasts 15 days or more ahead. Even so, centres tend to calibrate their models to perform better when storms threaten their own countries, but pay less attention to mid-ocean storms passing shipping lanes. None of the models can consistently produce accurate forecasts for tropical cyclones due to their complex physics and rapid development. Accuracy starts to deteriorate after 3-5 days, leading to even larger uncertainties between 5-7 days.
Most weather routing service providers use only one forecast centre. While the quality of forecasts may be sufficient for basic eather routing, their level of detail is often not sufficient for ship motion response prediction and voyage optimisation. This is particularly true in predicting the sea and swells generated by tropical cyclones. Currently, ocean weather forecast adopt a ‘man-machine’ mix, in which experienced forecasters quality-control the sea surface pressure forecasts, providing input to a marine boundary layer wind model, from which the output drives a numerical wave directional spectra model.
A ‘Super Ensemble’ forecast approach takes the best of each national centre’s forecast and quantifies the uncertainties in wind and waves. At each grid point in every forecast horizon extending to beyond 15 days, the values for mean and standard deviation can be used to judge the accuracy levels of the forecast. Such consensus knowledge will provide greater insight in selecting the optimum route, taking into account the risk of heavy weather damage as well as predicted fuel consumption for virtual arrival due to forecast uncertainties. Besides more accurate wind and wave forecasts, voyage optimisation takes into account sea surface currents, enhanced by satellite measurements, since they can significantly impact ship speed and fuel consumption.
The benefits of voyage optimisation can be further extended to ship design, fleet deployment, and operational logistics. In this regard, the technology is able to:
- Determine ship design criteria such as speed, sea margin, maximum ship motions, and bending moment by repeatedly simulating voyages using historic wind and wave hindcast databases;
- Optimise the deployment and schedule of vessels for the trade route taking into consideration schedule reliability, fuel cost, and seakeeping capability;
- Estimate the probability of on-time arrival so that shore-side operations can be efficiently scheduled;
- Extend the fatigue life of ship structures by predicting stress cycles and providing onboard seakeeping guidance to reduce ship stresses.
LATEST PRESS RELEASES
Research has shown that 80% of all hydraulic failures can be directly attributed to oil contaminatio... Read more
Turku, Finland, 09.10.2018. Auramarine, the leading provider of engine room fuel supply and auxiliar... Read more
Geislinger GmbH, located in Salzburg, Austria, celebrates its 60th anniversary this year which coinc... Read more
Expanding its global presence, MyTaskit has signed its first international reseller. Auckland, New Z... Read more
Vickers Oils are pleased to welcome Ian Bower, Business Development Manager and Esther Murray, Marke... Read more