How Does Uncertainty About Expected ESD Savings Affect the Investment Decision?
Uncertainty about the expected saving from an Energy Saving Device (ESD) affects the investment decision when the actual fuel reduction may differ so substantially from the prediction that the economic outcome changes. A Pre-Swirl Stator (PSS), Pre-Duct, Propeller Boss Cap Fins (PBCF) or Rudder Bulb is assessed using ship data, hydrodynamic calculations and expected operating conditions. There is uncertainty about the calculated power improvement, its translation into fuel consumption and the conditions in which the ship will actually use the saving. The significance of this depends on the financial margin between the expected return and the minimum saving required. If a retrofit also delivers sufficient benefit under less favourable, technically plausible conditions, further refinement of the prediction need not be decisive. If economic feasibility is close to that threshold, a targeted measurement, additional validation or a different design may change the investment decision. The issue is therefore not only how much uncertainty exists, but above all which uncertainty actually affects the decision.
What Uncertainties Underlie a Predicted ESD Saving?
A predicted ESD saving results from a comparison of propulsion performance with and without the proposed device. Both configurations are assessed using the available ship geometry, expected flow and relevant operating conditions. The reliability of the predicted saving therefore depends on both the input data and the calculation method used.
For a retrofit, the actual stern shape may, for example, differ from the original drawings. The existing propeller geometry and distance to the rudder must also be known with sufficient accuracy. If a Pre-Swirl Stator is designed for an incorrectly predicted wake field, the chosen fin setting may match the actual inflow to the propeller less well. For a PBCF, uncertainties about hub geometry and local vortical flow may instead be important.
Every hydrodynamic calculation also has limitations. Numerical verification examines factors including the effects of mesh refinement and convergence, while validation examines whether the model prediction agrees sufficiently with experimental data. A Computational Fluid Dynamics (CFD) model may predict overall speed–power performance reasonably well without the individual ESD contribution having been established with sufficient accuracy. The investment concerns the difference between the two configurations, rather than solely the ship’s absolute power demand.
The relationship between the predictions is also important. If the same calculation method is used for the configurations with and without an ESD, certain model deviations may occur in both results. Some deviations may partially cancel each other out when the power difference is determined, while others may distort the calculated ESD effect. The accuracy of the two absolute power predictions is therefore not necessarily the same as the accuracy of the predicted saving.
Extrapolating model tests to full scale may also add uncertainty. The boundary layer and local flow around the stern at model scale are not entirely the same as those of the actual ship. ESDs that depend strongly on the inflow to the propeller in particular may therefore perform differently from what would be expected from the model test alone. Original experimental research into various ESD principles has shown why scale effects and the individual forces on propulsion components must be assessed carefully.
Even when the hydrodynamic prediction is well substantiated, the annual fuel saving remains dependent on energy conversion and operation. A lower power demand may change the main engine’s specific fuel consumption. A different operational profile may also mean that the ship spends fewer hours at the speeds and propeller loadings investigated.
The latter is a different type of uncertainty from an insufficiently validated CFD model. The calculated ESD performance may be technically correct, while the annual return falls short because the ship is used differently. In the investment assessment, it must therefore remain clear which uncertainty comes from the performance prediction, which from future operating conditions and which from the financial assumptions.
Why Is the Highest Predicted Saving Percentage an Insufficient Basis for a Decision?
The highest calculated ESD saving often occurs at a particular operating point. That need not be the speed, draught or propeller loading at which the ship operates for most of the year. A design with the highest peak saving therefore does not necessarily deliver the greatest annual fuel reduction.
For an investment decision, the expected saving must be linked to the operating conditions in which it can actually be used. A favourable result at a nominal cruising speed carries less economic weight if the ship spends few hours there. Conversely, a more modest improvement at frequently used speeds may be important over a full operating year.
The reliability of the prediction is another consideration. Two designs may show similar annual fuel savings, while more suitable validation data are available for one than for the other. That difference is relevant to the confidence with which the expected return can be assessed.
The importance of this is evident from published comparative CFD research into the same Pre-Swirl Duct configuration. Different research teams predicted varying power differences, including both savings and deteriorations. The prediction of the wake field and the way in which the propeller was modelled particularly affected the results. This does not mean that CFD is unsuitable as a method, but it does mean that a small predicted ESD saving is not automatically accurate enough to support an investment decision.
What a reported uncertainty margin represents must also be clear. A spread in results from different CFD settings or research teams shows how sensitive the prediction is to certain modelling choices. It does not automatically follow from this how likely the actual saving is to fall within that spread. A separate, sufficiently substantiated uncertainty analysis is needed for that.
The same caution applies when comparing designs. If the calculated performance difference is small relative to the uncertainty of the comparison, the ranking may be insufficiently certain. The design with the highest calculated value need not then be demonstrably better.
A difference in prediction confidence also does not automatically mean that the design with the least uncertainty is financially preferable. Under certain circumstances, a higher expected return may justify greater risk. To assess that trade-off, it must be clear how the different possible outcomes affect the business case.
For a technical manager, the stated saving is therefore not the only relevant consideration. The assessment must also show the operating conditions to which the result applies, how the annual benefit has been determined and how reliably the difference between the configurations investigated has been substantiated. Only then does the saving percentage become sufficiently meaningful for an investment decision.
How Does the Minimum Saving Required Determine Investment Risk?
The significance of performance uncertainty depends strongly on the saving needed to justify an ESD investment economically. A possible deviation from the prediction becomes particularly important if it means that the minimum financial return required is no longer achieved.
That threshold is determined for the specific project. The additional investment, fuel price, relevant operating hours, payback objective, remaining operating period and any other net benefits together determine the annual saving needed. For charter operation, the proportion of the financial benefit that actually accrues to the party bearing the investment must also be established.
For an assessment with an unchanged operational profile, the minimum annual fuel saving required can be translated into a percentage of the relevant fuel consumption. A practical comparison thus emerges between the expected ESD performance and the performance needed to meet the financial objective.
A hypothetical example illustrates this decision without using investment amounts. Suppose a retrofit must deliver a minimum annual fuel saving of 2.5 percent under the chosen financial assumptions. The technical assessment predicts 3 percent, but a separately investigated, less favourable scenario produces 2 percent.
The predicted saving is then half a percentage point above the required threshold, while the less favourable outcome remains half a percentage point below it. With the same relevant annual fuel basis, that scenario represents a third less saving than the original prediction. The investment therefore does not meet the chosen financial objective under all the conditions investigated.
This does not automatically mean that the retrofit must be rejected. The credibility of the different outcomes, the consequences of disappointing performance and the uncertainty the investor is willing to bear are also relevant. Without reliable probability information, neither an expected financial return nor a probability of payback can be calculated from the two scenarios.
All percentages in this example are hypothetical. The minimum saving required is solely an illustrative internal decision threshold, rather than a general performance requirement for ESDs. Nor does the less favourable scenario constitute a statistically established lower bound.
If a conservative, technically substantiated scenario still produces a sufficient return, the investment may be less sensitive to uncertainty about the exact saving percentage. If feasibility is close to the required threshold, however, the quality of the prediction becomes more important.
The decision threshold primarily helps to focus the investigation on which uncertainty must be reduced. If the greatest risk arises from insufficiently reliable inflow to the propeller, additional hydrodynamic validation may be worthwhile. If it is mainly the future number of sailing hours that is uncertain, further investigation of the operational profile may have greater value.
Where operating conditions change substantially, a single minimum saving percentage is not always sufficient. It is then preferable to compare the full annual net returns or the net present value directly with the financial objective. In this way, different operational profiles or cash flow distributions are prevented from being incorrectly assessed against the same percentage threshold.
Why Can a More Robust ESD Design Be More Attractive than Higher Peak Efficiency?
A robust ESD design retains enough of the intended hydrodynamic benefit under relevant variations in operating conditions. This does not mean that the saving must remain the same at every speed or draught, but that the design is not solely dependent on a narrowly defined favourable operating point.
A Pre-Swirl Stator may, for example, be closely matched to the inflow at a particular cruising speed. If a ship regularly switches between different loading conditions or consistently operates more slowly than originally envisaged, performance under those other conditions may become more important to the annual return.
An alternative stator geometry may have a slightly lower peak saving, but deliver a more favourable or more consistent result across the expected operational profile. The investment benefit then arises not because the device is technically simpler, but because performance better matches the economically important conditions.
Sensitivity to the actual geometry may also be relevant. A design whose performance changes substantially with small deviations in position or profile shape places greater demands on the available ship data and installation. If that accuracy cannot be sufficiently assured, a less sensitive alternative may be more attractive.
A distinction must be made here between technical robustness and prediction confidence. An ESD may have low sensitivity to speed variations while the absolute magnitude of the saving is still insufficiently validated. Conversely, a thoroughly investigated device may demonstrably perform well within a limited operating range but offer little benefit outside it.
Structural suitability also remains a separate condition. A geometry with low hydrodynamic sensitivity to changing operating conditions is not thereby automatically strong enough or resistant to cavitation and vibration. Such risks must be assessed using appropriate technical methods.
A robust design is therefore not automatically better proven or structurally safer. These characteristics must be assessed separately and then linked to the expected operation. This also means that a more highly optimized ESD must not be rejected in advance. If the greater benefit is substantiated with sufficient reliability and the ship operates within the relevant range for a long period, that design may remain economically more attractive.
The investment choice therefore follows from the combination of expected annual performance, sensitivity to relevant changes and confidence in the calculated results. Higher peak efficiency is decisive only when the benefit is also sufficiently plausible under the conditions that actually matter.
How Are Technical and Commercial Uncertainties Assessed Together?
An ESD may fully achieve the predicted hydrodynamic improvement while the financial return still disappoints. This may happen, for example, if the fuel price falls, the ship sails fewer hours or the investor receives only part of the saving achieved.
Technical and commercial uncertainties must therefore remain separately identifiable, but must be incorporated together in the final business case. A change in the operational profile shows why this matters. If a bulk carrier starts consistently sailing more slowly, both propeller loading and operating hours at particular speeds may change. Together, those changes affect the annual fuel saving. It would be incorrect to calculate the effect on the operational profile and then apply the same change again as a separate correction to ESD performance.
The relationship between shaft power and fuel consumption also needs a consistent treatment. Uncertainty in the calculated power saving affects the engine’s fuel demand. How specific fuel consumption changes at the new operating point may also be uncertain. The results must be combined through the relevant power and engine relationships, rather than by adding separate percentage uncertainties without further substantiation.
Fuel prices are another uncertainty. They change the financial value of a quantity of fuel saved, but not the hydrodynamic performance of the ESD. In the assessment, it must therefore remain clear whether a disappointing financial outcome is caused by a lower achieved fuel saving or by a lower value per tonne saved.
Avoided emission costs may also depend on future prices, the scope of the regulations and the contractual allocation of costs. Such benefits may be included only insofar as the saving concerned actually generates a financial cash flow for the investor. The same avoided costs must not be counted several times under different names.
Under a time charter, the owner may invest while the charterer pays the fuel costs. Financial uncertainty for the owner may then lie mainly in the arrangements for compensation or future charter revenue, even when the technical ESD performance is well substantiated.
Consistent scenarios are therefore developed to support a reliable decision. The operational profile, hydrodynamic performance, engine consumption and commercial conditions must fit together within them. A less favourable scenario may, for example, consist of a credible combination of fewer relevant sailing hours and a lower fuel price, but need not automatically use the worst conceivable assumption for every component.
The scenarios must show the realistic conditions under which the financial objective of the investment is or is not achieved. Systematically combining all optimistic or all pessimistic assumptions may give a misleading picture if those conditions are difficult to reconcile with one another.
Without sufficient data, the different scenarios must also not be presented as equally likely. A base scenario, a less favourable scenario and a more favourable scenario are primarily tools for investigating the sensitivity of the investment decision. They do not automatically form a statistical probability distribution.
When Does Additional Investigation Provide Sufficient Value for the Decision?
Additional investigation is especially worthwhile when its outcome may change the investment decision. A more accurate ESD prediction has little economic significance if the remaining uncertainty does not materially affect the choice between the available alternatives.
Which uncertainty constrains the investment decision must therefore be established first. If the existing propeller geometry is insufficiently known, a targeted measurement or 3D scan may be needed. If the local flow is not predicted reliably enough, further CFD validation or a model test may be worthwhile. If the future operational profile is unclear, analysis of operational data may instead provide greater value.
These investigations provide different kinds of information. A 3D scan makes the geometric baseline more reliable, but does not thereby validate the predicted power saving. A model test may support the hydrodynamic prediction, but does not remove uncertainty about future operating hours.
The choice of additional investigation must therefore be linked to the minimum saving required. If a project remains clearly above the financial decision threshold, even under credible less favourable assumptions, further investigation may add little value. If a relatively small deviation takes the outcome below that threshold, a targeted study may be decisive.
Whether the investigation can distinguish sufficiently between outcomes is important here. A measurement campaign with uncertainty comparable to the performance difference that is decisive for the investment may not resolve the decision. An additional CFD calculation likewise has little value if the dominant uncertainty comes from future operating hours or contractual cash flows.
The expected value of information is determined by the possibility that better data will lead to a better choice. This may mean that an unfavourable investment is prevented, a promising project proceeds after all or a more suitable design is selected. Without reliable probability distributions, this value cannot responsibly be calculated as an exact expected monetary amount.
The costs and remaining limitations of the investigation also count. An additional test is economically worthwhile only if the potential improvement in the decision outweighs the investigative effort and any implications for implementation. More information does not automatically mean more value for the decision.
Project timing plays a particular role. If an ESD can be installed during a planned docking, postponement may affect the next available installation opportunity. Additional investigation may lead to a better-substantiated choice, but also to lost years of savings or a later separate docking. These consequences must be taken into account when determining how much further investigation is worthwhile.
The best investigation plan is therefore not necessarily the most extensive. It focuses on the uncertainty that can actually change the decision and provides the necessary information before the investment or installation opportunity has passed.
How Can Performance Agreements Limit Uncertainty and Financial Risks?
Technical performance predictions and contractual arrangements serve different functions. A hydrodynamic calculation indicates the improvement expected under certain conditions. A performance agreement determines which outcome is contractually relevant, how it is assessed and what consequences a deviation may have.
A predicted saving percentage is not automatically a guarantee. If a performance guarantee is agreed, its scope must be clear. A reduction in required shaft power at a particular speed, for example, is not the same as a reduction in annual fuel consumption.
The reference must also be fixed. If an ESD is installed during docking, the measured improvement may partly result from hull cleaning, propeller maintenance or other work. If an agreement concerns only the ESD contribution, it must therefore be determined how the situation without the device is established.
For a useful performance assessment, the relevant ship conditions, measurement method and treatment of uncertainty must also have been agreed. If a measured saving is smaller than predicted, it must be clear how the presence of a material deviation is assessed and which follow-up steps are possible.
A contractual threshold has a different meaning from the accuracy of a measurement method. An agreed minimum performance may be commercially decisive, but the measurements used must be reliable enough to assess whether that threshold has actually been met. Otherwise, the assessment of the performance agreement itself may create new uncertainty.
Such an agreement may, for example, concern additional analysis, a check of the installed geometry or the treatment of an established performance deviation. The rights and obligations arising from it depend on the specific contract.
A specific model addendum for ESD retrofits during time charter has been available since 2026. It provides the owner and charterer with a framework for agreeing retrofit costs, timing implications, and operational and environmental benefits. The technical performance assessment is developed in a separate Performance Assessment Protocol.
This addendum does not, however, automatically guarantee that an ESD delivers a particular fuel saving. Nor does it determine, without further agreement, what performance the supplier must deliver. The relationship between owner and charterer must be distinguished from any arrangements between the buyer and the ESD supplier.
Contractual clarity can distribute the financial consequences of disappointing performance more effectively, but does not change the ship’s actual hydrodynamic performance. It must also be assessed whether agreed compensation can actually be recovered and whether the party taking on the risk can meet its contractual obligations.
For an investment decision, both a sufficiently reliable technical prediction and workable contractual arrangements are therefore needed. An agreement can distribute uncertainty more effectively, but does not replace the technical substantiation.
When Is It Justifiable to Invest, Postpone or Decide against Proceeding?
The final decision depends on the expected ESD performance, the reliability of its substantiation and the consequences if the saving falls short. It must first be established that the proposed device is structurally and operationally suitable for the ship. Uncertainties about strength, necessary clearances or relevant cavitation and vibration risks cannot be accepted solely as financial risks.
Once technical suitability has been sufficiently demonstrated, the economic decision can be made. Investing may be justifiable if the expected net returns are attractive and the business case remains sufficiently robust under credible less favourable conditions. Absolute certainty about the eventual saving percentage is not required.
The financial assessment must, however, reflect the remaining operating period. A retrofit whose cost is recouped only after the expected sale or a planned major modernization requires different assumptions from a device that can continue to function in the same configuration for a long period. A possible residual value or future charter compensation may be included only if there is a sufficiently substantiated expectation of it.
Postponement may be worthwhile if the outcome is close to the financial decision threshold and a targeted investigation can provide enough information to change the choice. It may also be sensible to assess an ESD together with a planned propeller replacement or other propulsion modernization. The potential benefits of further substantiation must then be weighed against the costs and consequences of waiting.
A different design may be preferable if the original alternative has a higher predicted peak saving, but its expected annual performance is too sensitive to conditions that vary substantially during operation. It must then be established that the alternative actually produces a more favourable or more reliable result under the relevant conditions.
Deciding against proceeding may be justifiable if even a favourable but technically credible scenario offers an insufficient economic return, if the expected saving cannot be substantiated with sufficient confidence, or if the remaining operating period is too short to make appropriate use of the investment. A lack of workable agreements on the allocation of financial benefits may also be a decisive commercial objection.
For a substantiated choice, it must therefore be clear what annual performance is expected, which uncertainties are material and how the outcome compares with the minimum return required. It must also be known whether additional investigation can actually improve the decision and what consequences postponement has for the installation schedule.
Uncertainty about an ESD saving is therefore not solely a technical disadvantage that must be reduced as far as possible. The issue is which uncertainty can change the economic decision. An investment may be justifiable if the expected value holds up under a sufficiently broad range of credible conditions. If the outcome is close to the decision threshold, the quality of the substantiation and the choice of additional investigation become decisive. An investment decision thus emerges that relies not only on the highest predicted efficiency, but on a sufficiently reliable assessment of the performance the ship can actually use.
This Article within the Series
A deviation from the predicted fuel saving directly affects the payback period of an ESD retrofit, together with the fuel price, operating costs and available operating hours. If the financial outcome depends strongly on uncertain assumptions, particular importance attaches to whether a feasibility analysis provides a sufficient basis for the investment. Both considerations form part of the economic assessment of ESD investments.