Seaways – May 2018

10   | Seaways | May 2018 Read Seaways online at http://www.nautinst.org/seaways

Feature: DUKCM – A note of caution

Dynamic under-keel clearance management promises greater efficiency, but it is

essential to make a significant allowance for potential errors

DUKCM – A note of caution

Tim Hallpike

MSc CMarSci MIMarEST MNI

A

n article by Captain Jonathan Pearce (Seaways, March 2018)

advocates the use of real-time environmental measurements

combined with a suitable computer-based programme to

determine the under-keel clearance. While this dynamic

approach is clearly more efficient than the traditional ‘static rule’

approach (typically 10% of the draught), it assumes that real-time

measurements are always accurate. Unfortunately, this is rarely the

case.

The degree of accuracy depends on the inherent accuracy of

the sensor/equipment, coupled with the measures adopted to both

calibrate the sensor/equipment and validate the readings. Even then,

there will still be random erroneous readings and, for each parameter

being measured, there will be a standard error (SE). Since some of the

parameters used to determine the dynamic under-keel clearance are

based on calculation rather than measurement, eg squat, these values

will also be subject to error, especially when there is more than one

possible method of calculation for these values1,2.

Creating an error budget

Whenever there is more than one source of error, the widely accepted

method of determining the overall error is to compile an ‘error budget’

• sometimes referred to as an ‘uncertainty budget’. A suggested ‘error
  budget’ for dynamic under-keel clearance is as follows:

The above 2σ error values are considered to be conservative

estimates. It only requires one of the larger sources of error to increase

by a small amount, and the RMS value will increase significantly.

Factors affecting the error budget

CHARTED DEPTH ACCURACY

It should be noted that this level of accuracy is typical of the most

accurate (‘special order’) hydrographic surveys. The error could well

be larger, depending on the rate of siltation and survey/dredging

frequency. For charts using zones of confidence (ZOC) to denote

depth accuracy and seabed coverage, the highest level of depth

accuracy is only + 0.5m + 1% of the depth (ZOC A1).

SQUAT CALCULATIONS

It is very rare for commercial vessels to conduct squat trials. Instead,

the required squat curves are generated using one of the numerous

formulae currently available1,2. Each formula generates a slightly

different squat value, some differing quite significantly from others,

especially when the depth-to-draught ratio (h/T) is <1.2^3.

DRAUGHT CALCULATIONS

The draught forwarded by the ship to the port authority prior to arrival

is usually based on a calculation and is a function of how much fuel

and fresh water has been used since the last port. It is therefore subject

to error and, while it is fairly standard practice for the Pilot to check

the ship’s draught before boarding, the accuracy will still be suspect,

especially in adverse conditions.

HEIGHT OF TIDE MEASUREMENTS

Most port authorities rely on tide gauges for tidal information. The

accuracy of these gauges depends on a number of factors, according

on the type of sensor being used. For example, in order to obtain

an accurate tidal reading from a pressure sensor, it is necessary to

determine the mean density of the water column above the sensor – a

value that is likely to change significantly in estuarine waters with large

tidal ranges. In addition, tide gauges should be calibrated at regular

intervals. Ideally, there should be a minimum of three sensors in order

to detect/dismiss a defective sensor. Where the approaches to the port

are long, there may well be a need to establish a network of tide gauges

in order to produce a co-tidal chart. Co-tidal corrections can be quite

substantial.

WAVE HEIGHT AND PERIOD

The measurement of wave/swell height and period is usually carried

out by means of a buoy-mounted sensor. There are two types of sensor

in common use:

l Linear accelerometers

l Real time kinematic (RTK) GPS receivers.

Both types of sensor can achieve +3–5cm accuracy. In both cases,

however, sensors capable of this level of accuracy are expensive – to say

nothing of the cost of purchasing, installing and maintaining the buoy

and a reliable data link.

RTK GPS cannot be relied upon to provide continuous readings for

Sources of error

Estimated error values (metres)

SE (2σ) SE^2

Charted depth accuracy ±0.3 0.

Squat calculation ±0.4 0.

Draught calculation ±0.1 0.

Height of tide measurement ±0.1 0.

Wave/swell height measurement ±0.1 0.

Wave/swell-related pitching

resonance

±0.2 0.

Heel of vessel (turning, wind,

rolling resonance)

±0.4 0.

Seawater density variation in

estuarine waters

±0.1 0.

Sum of the SE^2 values 0.

Root mean square (RMS) value

(2σ)

± 0.

Dynamic under-keel clearance – error (uncertainty) budget

DUKCM_SGS, author.indd 10 18/04/2018 13: