What it takes to have a win-win-win outcome for shipowners, charterers, and the environment

Outdated Charter Party Agreements do not incentivize efficient ship design or operations

More than half of the world’s shipping fleet is under charter in one form or another. A Charter Party agreement (CP Agreement) is a maritime contract and performance warranty between a charterer and a shipowner whose ships are hired to transport cargo or passengers.??The contract clauses have evolved over the past century, with the most recent update by BIMCO to address the new CII regulations set by IMO starting 2023. Traditionally, due to market competition and unpredictable downtime because of weather or ship operations, charterers have insisted on using “checks and balances” to penalize underperforming ships using speed and fuel claims when calm weather speed is below or daily fuel consumption is above the agreed values. Shipowners, on the other hand, try their best to get business by over-promising their ship’s performance in a time charter, using all means to avoid penalties; or to profit from demurrage charges by sailing at maximum speed in a voyage charter. This adversarial relationship does not build trust, and often leads to inefficient and unsafe voyages with heavy weather damages.?Many benefits of fuel-saving measures, such as improved antifouling systems, hull/propeller polishing, and voyage optimization for Just-In-Time Arrival (JIT Arrival), etc., unfortunately are not realized due to conflicts of interest.?Worst of all, the unintended consequence of wasted fuel leads to increased GHG emissions, further contributing to global warming.

“Checks and balances” becomes a “Cat and Mouse” game

Weather routing was the first attempt to carry out this “checks and balances” under Charter Party.?Service providers first started to utilize wind and wave forecasts from the US Navy’s Fleet Numerical Oceanography Center (FNOC) to advise charters and avoid heavy weather in the early 1950s.?Besides routing ships away from bad weather, service providers also conducted post voyage analysis to verify ship locations, and Beaufort Sea state numbers in shipowners’ noon reports as a basis for fuel and speed claims.?The advent of satellite AIS and supercomputers running global numerical weather forecast models have made “creative” noon reports more difficult to get away with, but also keep the arbitration lawyers busy. The points of contention are often centered on the severity of observed ocean conditions and vaguely-defined “about clauses” in warranted calm-weather speed and fuel consumption.

To avoid claims, a ship captain would simply increase power to maintain calm-water speed, and shift the additional fuel consumption to bad-weather days, when the ship can slow down without being penalized.?This practice precludes tactical routing advice such as slowing down in good weather for the storm to pass instead of fighting with bad weather in order to make up the fuel difference. The CII loophole also incentivizes the captain to take a longer route in calm weather to improve CII rating while disregarding fuel wastage and associated GHG emissions. ?Similarly, shipowners are reluctant to adopt any win-win optimization strategy such as Just-in-Time Arrival (JIT arrival) or Virtual Arrival in order to profit from demurrage charges in a voyage charter; or fear of speed/fuel claims in a time charter.

Effect of uncertainties in long-range weather forecasts on a ship’s speed profile

Another consideration is the uncertainties in wind and wave forecasts for ships on long voyages. While major weather centers around the world produce long-range forecasts beyond 15 days, acceptable accuracy for wind and wave conditions around major storms is normally up to 5 days, and even shorter in tropical storms for ship routing purpose. This is due to the inherently chaotic nature of the earth's atmosphere and the imperfect depiction of its initial state in a non-perfect numerical forecast model. An omitted small disturbance somewhere on land can develop into a major storm several days later. Unexpected changes in the storm track and intensity can drastically affect the safety and predicted fuel consumption of a voyage, not to mention its planned ETA. ?Ship captains naturally will take a conservative route strategy, including “hurry up and wait”, trying to avoid charter party claims while ensuring on-time arrival.

Most weather routing companies rely on one source of deterministic current, wind, and wave forecasts as the input to their route optimization algorithms, with updates when the next forecast becomes available. Without quantifying the uncertainties or risk of heavy weather along the route later in the voyage, both shipowners and charterers need to assume that the forecast, ship model and optimization algorithms are all “perfect” in order to minimize fuel consumption while arriving on time. In reality, much of the savings touted in case studies could not be consistently demonstrated due to unpredictable weather as well as modeling errors.

Prediction uncertainties can be quantified by voyage simulations using ensemble weather forecast

Simulations based on ensemble forecasts (e.g. 31-member GEFS from NOAA or 51 members from ECMWF) can generate statistics (e.g., mean, standard deviation. and histogram) of weather, ship motion, fuel consumption, and ETA along the selected route.?The decision-makers can then select the optimal route and speed profile based on specified response threshold limits for safety and other voyage KPIs, including unscheduled maintenance as well as charter party claims with quantified risk and uncertainties.

An ETA with quantified uncertainty may eventually change the “first-come, first-served” port policy that is a century-old tradition, when long-range forecasts were not available, and ETAs could not be accurately predicted. Consequently, ships that arrive early may have to wait for days for cargo unloading/loading. Perhaps a “Virtual Arrival into a virtual queue”, with penalties for delay, could change the motivation for “hurry up and wait”, which can result in significant increases in fuel consumption and GHG emissions.?Simulations using ensemble forecasts have the potential to reduce wasted fuel by providing real-time decision support for JIT arrival, with known uncertainties transparent to all parties.

Ship performance models need to be improved

If the objective is only to avoid bad weather, a simple speed reduction curve that predicts the percent reduction of calm weather speed as a function of sea state or Beaufort number in head, beam, and following seas is sufficient. The ship is assumed to be sailing under constant power, and total fuel consumption is essentially based on voyage time.

On the other hand, if the objective is to minimize total fuel cost and CII for a desired arrival time, prediction of engine power, RPM, and fuel consumption for different ship drafts and speeds in all weather conditions, becomes necessary. This requires modeling the entire hull, engine, and propulsion system. Even more sophisticated modeling is required when motion and seakeeping responses in heavy weather are part of the voyage optimization algorithm. In this case, wave directional spectra or parameterized seas and swells in the wave forecast are needed to predict roll, pitch, accelerations, slamming, and propeller racing. Simply using significant wave height or sea state, and ignoring the effect of swells from different directions, will not provide accurate prediction of ship responses.?Empirical models and analytical tools have been used in the past, but may have reached their limits due to the complex wind shielding of ship superstructure geometry, as well as the dynamic interactions between motions and propulsive efficiencies in waves.?Regression analysis based on small numbers of questionable noon reports simply are not reliable for predicting required horsepower in different drafts and infinite combinations of wind, wave and current conditions.?

Can high frequency ship data and AI help in creating a ship’s performance Digital Twin?

Recent advances in Artificial Intelligence (AI), specifically Artificial Neural Network (ANN), has inspired many data scientists in startup companies try to solve the modeling problem with or without high frequency data.?ANN is a form of nonlinear function approximation. However, modelers should be cautioned not to treat the problem as an unsupervised “black box” input and output without understanding the physics. High frequency ship data, while proving to be an invaluable source for AI machine learning, could also give a false impression on goodness of fit, when data validation sets and training sets are drawn from the same population.

There is a major difference between supervised and unsupervised learning of a neural network. In supervised deep learning, the model is based on the principles of naval architecture. AI is used to train the model coefficients of the motion equations, as well as forces and moments exerted on the ship by its controls and environment. Whereas, unsupervised learning treats the model as a “black box” characterized by observed inputs and outputs. Training usually takes a lot longer, with higher residual mean square error. Even with a well-trained neural network model, out-of-kilter predictions are often encountered when input data were not in the training set or not constrained by physics. A third party is therefore needed to validate the accuracy of these models (Digital Twin) for the intended applications.

Trust and transparency are the keys to achieve a win-win-win outcome

A modernized charter party agreement that utilizes the ensemble wind and wave forecast plus a performance Digital Twin in voyage optimization could create a win-win-win scenario for charterers and shipowners by saving fuel for just-in-time arrival.?Efficient and safe voyages will lead to substantial reductions in GHG emissions, not to mention avoiding expensive legal costs to settle claim disputes.

In order to achieve these objectives, charterers and shipowners must work together and build trust. Trust comes from transparency. Shipowners can put their ship performance models on an open platform that service providers can use to simulate proposed voyages and estimate ETA, fuel consumption, and CII.?Classification societies equipped with intimate technical knowledge of the ship’s history can Type Approve (validate) the performance model (Digital Twin) for the intended applications.?Charterers can confidently choose the vessel best fit for their purpose, not only in terms of chartering rate, but also expected fuel cost and GHG emissions, consistent with the charterer’s publicly-stated ESG goals. Shipowners can also use the Digital Twins to plan future fleet maintenance and renewal for better CII ratings. ?A better-performing ship equipped with advanced energy-saving devices and/or a better hull/engine maintenance history can differentiate itself and command a higher rate premium.

Service providers can access the ship performance models to run voyage optimization utilizing ensemble forecasts for JIT arrival, Virtual Arrival or any other routing strategy to increase safety and efficiency. Ship captains do not need to risk heavy-weather damage or waste time in creative fuel accounting to avoid speed/fuel claims. Charterers and owners may share the benefits of these strategies in a win-win-win relationship.

Working with BIMCO, the proposed charter party agreement would get rid of performance warranty clauses and related speed/fuel claims, but would include the following elements:

• ?A Type-Approved performance Digital Twin on an open platform under its IMO/MMSI number
• Service providers allowed to access the ship performance model to perform voyage simulation and optimization at the request of charterers or shipowners/managers
• A real-time monitoring system to relay shipboard high frequency performance data to shore-side cloud storage on?hourly basis
• Require updating of the Digital Twin on an agreed interval

The Type Approval process would ensure:

• An accurate and up-to-date unbiased ship performance model (Digital Twin) that can predict engine power, RPM, and fuel consumption at different ship speeds and drafts typical for the specific ship types, in all weather conditions
• Validation of the wind resistance, wave resistance, and form resistance results with past records (if available) or extrapolation based on naval architecture principles
• The latest engine-specific fuel consumption curves reflecting the engine running conditions and accuracy of monitoring instrumentation (fuel flow meter, torque meter etc.)

Open source Digital Twins will lay the groundwork for future autonomous navigation

The new open source framework of Type-Approved Ship Performance Digital Twin will allow accurate speed/fuel consumption predictions in different environmental conditions, enabling the creation of an optimal route that minimizes fuel consumption and avoids heavy weather damage while maintaining on-time arrival.?Adding a Ship Maneuvering Digital Twin on the same open platform will enable fully automated navigation for autonomous ships. Such digital twin based on Recursive ANN has been successfully deployed on small unmanned vessels.

Adding power/rpm and rudder angles (already acquired by the VDR and for navigation/machinery control systems) to the compressed AIS sentences will allow prediction of future positions of nearby ships in real-time from past maneuvers.?Sophisticated algorithm with input from ECDIS can figure out the best course of actions taking into consideration of anti collision/grounding and other safe navigation concerns.

When such systems become mature, ship automation systems could take over issuing commands for RPM and rudder angle in a closed-loop system together with nearby ships.?Only then will we be able to achieve autonomous navigation that relegates navigators to “watch standing” only.

We must accelerate our Digitization effort

We are living in a 5-D world: Decarbonization, Decoupling, Derisking, and Disruption of supply chains, and Digitization of information systems. The first 4 Ds have already drastically changed the traditional shipping business environment. Increasing occurrence and severity of storms caused by global warming, and a shortage of experienced seagoing crew, might have caused many recent accidents involving lost of life, cargo damage, and pollution. We need to accelerate our Digitization efforts by building accurate ship Digital Twins and utilizing ensemble wind and wave forecasts in order to assist shipowners, managers, and operators to navigate and survive in these treacherous waters. We should not allow outdated charter party agreements to further waste nonrenewable resources and damage our environment; nor can we afford to muddle through by just delivering cargo from port to port without demonstrating high standards of safety, fuel efficiency, and commitment for Decarbonization to save the environment.

Disclaimer: This article was not generated by Chat-GPT. Please contact the author at [email protected] for further discussions and explore future technical/commercial opportunities.