Hull and propeller performance in service with ISO 19030
A ship leaves dry dock with a smooth hull and a polished propeller, and from that day its performance drifts. Slime, weed and barnacles roughen the hull, the coating wears, the propeller blades pick up fouling and damage. The ship then needs more power for the same speed, or makes less speed with the same power. On a ship burning 30 tonnes a day, a few per cent of fuel soon adds up to the price of a hull cleaning.
Measuring that drift is harder than it looks, because the weather, the draught, the water temperature and the depth all change the power at a given speed by as much as or more than the fouling does. ISO 19030 sets out how to compare in-service data with a reference. This guide shows the working steps: correct the data, fit the speed–power curve, read the speed loss and power increase, check them against the physics of roughness, and turn the penalty into fuel, money and a cleaning interval. The numbers come from the speed–power curve fit, the hull roughness penalty and the hull cleaning interval calculators, with the site's ISO 15016 correction calculators for the weather.
Why hull and propeller performance drifts
Roughness acts on the frictional part of the resistance, which for a full, slow ship is most of it. The standard way to quantify it is a roughness allowance ΔC_F added to the friction coefficient, as a function of the equivalent sand roughness k_s and the Reynolds number. ISO 15016:2025 (E.7, G.8) gives it in the Townsin form:
ΔC_F = 0.044 [(k_s / L)^(1/3) − 10 Re^(−1/3)] + 0.000125 (minimum 0)
ΔR = ½ ρ S V² [ΔC_F(k_s, fouled) − ΔC_F(k_s, clean)]
A newly applied antifouling coating is already rough: 150 µm is the ISO default when no roughness survey exists. Schultz (2007) assigned equivalent roughness values to fouling states: deteriorated coating or light slime 300 µm, heavy slime 600 µm, small calcareous fouling or weed 1 000 µm, medium calcareous fouling 3 000 µm, heavy calcareous fouling 10 000 µm. Fouling is not sand grain, so those equivalents carry about ±30 % uncertainty.
The propeller fouls too, and usually faster. Propeller roughness typically costs a further 2 to 6 % of power on top of the hull figures. In-service measurement sees hull and propeller together; it cannot separate them without extra information such as a recent propeller polish.
The speed–power curve and the reference
Over the service range a ship's power is well described by a power law, P = a Vⁿ, with n typically 3 to 4.5 for displacement ships. Performance is judged against a reference curve for the clean condition: the sea trial curve at the same draught, a model test curve corrected to full scale, or, as ISO 19030 also allows, a reference period of the ship's own service data. Two indicators then describe the change:
Performance value (ISO 19030-2):
PV = (V_measured − V_ref(P_measured)) / V_ref(P_measured) × 100 %
Power increase at constant speed:
ΔP = (P_fit(V) − P_ref(V)) / P_ref(V) × 100 %
The performance value takes each measured point, asks what speed the reference curve gives for the measured power, and compares. A negative value is a speed loss. The power increase is the alternative indicator, evaluated at a chosen speed. Both describe the same change: with n about 3.4, a 7 % power increase corresponds to roughly a 2 % speed loss.
What must be corrected before the data mean anything
The fitted curve compares hull and propeller condition only if everything else has been taken out. The speed–power calculator assumes its data are at one draught and trim and already corrected for wind, waves, current, water temperature and depth, following ISO 15016 and ISO 19030-2. In practice that means the following checks for every data point.
Speed through water, not over ground
Resistance depends on the speed relative to the water. Speed over ground includes the current: STW = SOG − current along the track. Use the speed log, or GPS speed corrected for a known current. The propeller slip calculator makes the same split and is a useful cross-check: a sudden change in apparent slip in steady weather often points to a speed input that is wrong rather than a hull that changed.
Wind
The wind resistance calculator follows ISO 15016:2025 Annex C. The added resistance is the wind force at the relative wind minus the still-air resistance the calm-water baseline already contains:
R_AA = ½ ρ_A C_AA(ψ) A_XV V_WR² − ½ ρ_A C_AA(0) A_XV V_G²
ΔP_wind = R_AA V_G / η_D
The coefficients come from the Fujiwara regression or wind-tunnel data. The relative wind must be at the 10 m reference height, and the power uses speed over ground, as the ISO 19030-2 Annex G wind correction does.
Waves
For short head waves with little ship motion the STAwave-1 calculator gives R_AW = (1/16) ρ g H_1/3² B √(B / L_BWL) (ISO 15016:2025 D.31), where L_BWL is the length from the bow to where the waterline reaches 95 % of the maximum breadth. It applies only to waves within ±45° of the bow and has no speed or period dependence. Long swell, large motions and following seas need the spectral method or model tests, and in service analysis such points are better filtered out than corrected.
Water temperature and density
Warm water is lighter and less viscous than the 15 °C, 1026 kg/m³ reference, so a ship in the tropics looks slightly better than it is. The water temperature correction follows ISO 15016:2025 Annex E. The effect is about ±1 % of power between 5 and 30 °C, small but systematic by season and route. The sign confuses everyone: in warm water R_AS is negative and the corrected power is higher than measured.
Shallow water
The shallow water calculator applies Raven's method from ISO 15016:2025 Annex G, which splits the effect into a viscous increase scaling with (T/h)^1.79 and a sinkage term. It needs the depth along the track, and the power must already be corrected for weather.
Draught and trim
No correction above brings a ballast point onto a loaded reference curve. Keep one draught and trim per analysis, with its own reference curve, and never fit a single curve through mixed loading conditions.
Worked example: a 180 m bulk carrier ten months out of dock
The ship: L_PP 180 m, B 32 m, T 11 m at the loaded draught, wetted surface 7 000 m², η_D 0.70 assumed for the corrections. The sea trial at this draught gave five reference points from 11 kn, 3 300 kW, to 15 kn, 9 472 kW.
Step 1: correct one point
Take one noon-report point: 13.6 kn through the water, 13.4 kn over ground, 7 836 kW measured. The relative wind is 24 kn from 20° on the bow, the significant wave height 1.0 m from ahead, the sea temperature 25 °C. For the wind the ship has L_OA 190 m, a transverse projected area A_XV of 600 m² and a lateral area A_LV of 1 200 m² (the other superstructure inputs at the calculator's defaults); the bow length L_BWL is 40 m. For corrections this small the 1 % between shaft and delivered power can be ignored.
- Wind. ρ_A = 1.225 kg/m³, C_AA(20°) = 0.926, C_AA(0) = 0.761. Wind force 51.9 kN minus still air 13.3 kN gives R_AA = 38.6 kN. ΔP = 38.6 kN × 6.894 m/s / 0.70 = 380 kW.
- Waves. ρ = 1023.2 kg/m³ at 25 °C, √(32/40) = 0.894. R_AW = 1/16 × 1023.2 × 9.807 × 1.0² × 32 × 0.894 = 17.9 kN. ΔP = 17.9 kN × 6.996 m/s / 0.70 = 179 kW.
- Temperature. The reference curve gives 6 788 kW at 13.6 kn, so R_Tref = 6 788 × 0.70 / 6.996 = 679.1 kN. The density term is −1.85 kN and the friction term −1.85 kN, R_AS = −3.70 kN, ΔP_AS = −37 kW.
- Corrected power. 7 836 − 380 − 179 − (−37) = 7 314 kW.
- Depth. The noon position shows deep water. Had the ship been in 35 m, Raven's method would have taken another 341 kW (4.7 %) off; even 60 m would have been worth 107 kW (1.5 %).
Left uncorrected, this point has a performance value of −4.1 %. Corrected, −2.2 %. The weather alone nearly doubled the apparent fouling.
Step 2: fit the corrected data
Eight corrected points between 11.0 and 15.3 kn go into the speed–power curve fit, which fits ln P = ln a + n ln V by least squares:
| Speed (kn) | Power (kW) | Reference speed at this power (kn) | PV (%) |
|---|---|---|---|
| 11.0 | 3 573 | 11.260 | −2.31 |
| 11.8 | 4 446 | 12.008 | −1.74 |
| 12.4 | 5 327 | 12.664 | −2.09 |
| 13.1 | 6 331 | 13.324 | −1.68 |
| 13.6 | 7 314 | 13.902 | −2.17 |
| 14.2 | 8 387 | 14.473 | −1.89 |
| 14.7 | 9 547 | 15.035 | −2.23 |
| 15.3 | 10 776 | 15.580 | −1.80 |
- The reference points fit P = 0.95 V^3.40 exactly.
- The service data fit P = 1.058 V^3.385, R² = 0.9995. The exponent is close to the reference one, a sign the data are at one draught and well corrected.
- At 14 kn: reference 0.95 × 14^3.4 = 7 491 kW, service 1.058 × 14^3.385 = 8 013 kW. Power increase 7.0 %.
- At 8 000 kW: reference speed (8 000 / 0.95)^(1/3.4) = 14.27 kn, service 13.99 kn. Speed loss 2.0 %.
- Mean performance value of the points: −2.0 %.
A quick cross-check with the Admiralty coefficient, taking Δ = 180 × 32 × 11 × 0.80 × 1.025 = 51 955 t: 510 on trial at 14 kn, 477 now. The trial value is 7 % higher, the same ratio as the powers.
Step 3: is the loss plausible?
The hull roughness penalty calculator with this ship at 14 kn, 15 °C, the reference 7 491 kW treated as delivered power, η_D 0.70 and a clean hull of 150 µm gives:
- light slime (300 µm): +2.8 % power, 0.9 % speed loss;
- heavy slime (600 µm): ΔC_F from 1.12 × 10⁻⁴ to 3.55 × 10⁻⁴, ΔR = 45.3 kN on a total of 728 kN, +6.2 % power, 2.0 % speed loss.
The measured 7.0 % sits at heavy slime on the hull alone, or lighter hull fouling plus some propeller roughness. That is the useful answer: the measurement is in the range the physics predicts, and a diver's inspection should find slime rather than hard fouling.
Step 4: what it costs
On a 5 000 nmi passage at 14 kn and SFOC 175 g/kWh, the fuel and voyage cost calculator gives 31.5 t/day of main-engine fuel at the reference power and 33.7 t/day at the fouled power: 468.2 t against 500.8 t for the passage, 32.6 t more, about 17 900 at 550 per tonne. That extra fuel is also extra CO₂ in the ship's CII.
Step 5: when to clean
A 7 % penalty ten months after a full clean is an average growth of 0.7 % a month. The hull cleaning interval calculator treats the penalty as growing linearly and minimises the average cost of fuel plus cleaning:
c₁ = F × d × (p + s C_F e) / 100 cost per month of a 1 % penalty
T* = √(2C / (c₁ r)) optimal interval
With F = 31.5 t/day, 22 sea days a month, fuel at 550, allowances at 70 per t CO₂ with half the emissions under the ETS and C_F 3.114: c₁ = 31.5 × 22 × (550 + 0.5 × 3.114 × 70) / 100 = 4 567 per percentage point per month. A cleaning costs 60 000 plus one day off-hire at 20 000, C = 80 000. Then T* = √(2 × 80 000 / (4 567 × 0.7)) = 7.1 months, with the penalty reaching 5.0 % before each cleaning. That costs 271 400 a year in penalty plus cleaning, against 287 800 for cleaning every 10 months, a saving of about 16 400 a year and 85 t less fuel wasted.
How to read the result
Watch the trend of the performance value over time, not one number. Within a cleaning cycle it should fall steadily; after a hull cleaning or propeller polish it should step back up, and the size of the step tells you what the cleaning recovered. A PV that does not recover after cleaning points to residual roughness or coating damage, which only recoating removes; the cleaning calculator carries it as a residual penalty p₀ that no interval can remove. A loss that rises far faster than the average often follows an idle period in warm water, when calcareous fouling settles.
The same drift is visible in the engine's propeller curve. The propeller curve calculator reports how heavy a measured point runs against the clean-hull curve and how much of the light running margin is used up, which matters before the loss becomes a torque limit.
Common mistakes
- Speed log errors. A log reading 1 % high moves the mean performance value in the example from −2.0 % to −1.0 % and the power increase at 14 kn from 7.0 % to 3.4 %. A drifting log looks exactly like a hull being cleaned or fouled. Calibrate it, and compare against GPS speed in slack water.
- Speed over ground used as speed through water. A half-knot current at 14 kn is a 3.6 % speed error, larger than most fouling signals.
- Mixed draughts. Ballast and laden points on one fit give a meaningless exponent. The calculator warns when n falls outside 2 to 5; take that as a sign of mixed data, units or weather.
- Too little data or too narrow a speed range. Three points from 13.1 to 14.2 kn give n = 3.48 instead of 3.385 and trigger the warning that a span under 10 % in speed leaves the exponent poorly determined. Evaluate inside the range of the data; extrapolated speeds are flagged.
- Getting the temperature sign backwards. In warm water the correction raises the power.
- Brake power against a shaft-power reference. Use the same power for data and reference, or the gearbox and shaft losses show up as fouling.
- Correcting points that should be filtered. STAwave-1 does not cover beam or following seas, and a large wind correction carries a large uncertainty. Dropping such points is better than correcting them.
Limits and the standards behind the method
The speed–power fit follows the performance value definition of ISO 19030-2:2016 and the power-law fitting of ISO 15016:2025 and ITTC 7.5-04-01-01.2. A power law is a local description, so the result holds inside the speed range of the data. The roughness penalty follows the 1978 ITTC performance prediction method and ISO 15016:2025 Annexes E and G, with the Schultz (2007) equivalents for fouling. It holds the propulsive efficiency constant, so the real penalty is 5 to 10 % larger than computed. The cleaning interval uses a replacement-interval model (Jardine and Tsang) with linear average growth, and leaves out port rules, capture requirements, coating damage and the IMO biofouling guidelines (MEPC.378(80)). The corrections themselves are those of ISO 15016:2025 Annexes C, D, E and G, each with its own range of validity.
Calculators used in this guide
- MaritimeSpeed–power curve fitFit P = aVⁿ to speed and power data and compare it with a reference curve: power increase, speed loss and the ISO 19030 performance value.
- MaritimeHull roughness and fouling penaltyAdded resistance, power, fuel and speed loss from hull roughness or fouling via the ITTC-57 line and the ISO 15016 roughness allowance.
- MaritimeOptimal hull cleaning intervalCleaning interval that minimises fuel penalty plus cleaning cost, from the fouling growth rate, fuel and ETS prices and the cost per cleaning.
- MaritimeWind resistance (ISO 15016)Added wind resistance and power from the relative wind by ISO 15016:2025 Annex C with the Fujiwara coefficients or wind-tunnel data.
- MaritimeAdded resistance in waves (STAwave-1)ISO 15016 STAwave-1 added resistance in head waves from beam, bow length and significant wave height, with the power increase at speed.
- MaritimeWater temperature and density correctionISO 15016 Annex E resistance and power correction for sea water temperature and density against the 15 °C / 1026 kg/m³ reference.
- MaritimeShallow water correction (Raven / Lackenby)ISO 15016:2025 Annex G (Raven) shallow-water power correction — viscous increase and sinkage — plus the legacy Lackenby speed correction.
- MaritimeFuel consumption and voyage costFuel burned, time at sea, bunker cost and CO₂ for a passage from distance, speed and engine power or daily consumption.
Results are estimates. They do not replace the governing code, class or flag requirements, or a check by a competent person. Report an error in this guide
More guides
- Cargo securing calculation to CSS Code Annex 13
- How to do a draft survey calculation, step by step
- How EU ETS and FuelEU Maritime costs are calculated
- How to calculate a ship's CII rating, step by step
- Ship resistance and powering: from hull to installed power
- How to calculate ship squat and under-keel clearance