SHIP SQUAT PART 1:ARE WE OUT OF OUR DEPTH?

February 1st, 2008 by JCB
Posted in Articles, Features | 7 Comments »

 

A tanker in a restricted channel. According to ship squat tables this tanker should be aground! Photo :JCB

All pilots are aware that, at speed, ships display a tendency to sit deeper in the water, a phenomenon officially referred to as “squat”. Despite their being generally aware of squat, most pilots have no respect the squat tables calculated for a particular ship because they frequently indicate that a passage using historically proven safe under keel clearance (UKC) parameters is mathematically impossible! In the majority of ports the UKC parameters were established sometime around the time of Noah and traditional pilot training has meant that rather than relying on mathematical tables, pilots gain an instinctive “feel” for the ship with vibration, high exhaust temperatures or a breaking quarter wash indicating that a vessel’s speed is too high for the existing water depth which usually results in the speed being reduced before a squat induced grounding occurs! A pilot’s knowledge and experience of their own district is therefore considered more reliable than the tables and consequently the number of groundings solely resulting from squat are almost non existent and the only case that I can identify as being totally attributable to squat is the QE2 leaving Massachusetts in august 1992.

So, in theory, ship squat can be disregarded as a serious problem for pilotage navigation but these days the “we’ll pull her back a bit Capt as we go over the shoal” approach is not considered best passage planning practice so, in order that we can factor in squat, it is increasingly necessary to have accurate data regarding the causes and effects of squat.

Squat does exist! This vessel is steaming at about 12.5kts. Draft 7m in water depth 16m. Photo : JCB

As mentioned previously, information provided by the ship’s squat tables rarely tallies with the pilot’s / port’s established passage planning speed / UKC guidelines but there are now two key reasons why pilots must take squat seriously. Firstly, should a pilot be unfortunate enough to be involved in any incident the passage plan and master / pilot exchange will be examined in detail and if a pilot hasn’t discussed the ship squat characteristics with the Master then he will be condemned by the investigators, regardless as to whether or not squat was of relevance to the incident. The second factor is that well trained “bridge teams” now utilise squat tables when calculating safe UKC parameters in their passage plans so the Master / pilot relationship can get off to a frosty start if the pilot breezes up to the bridge and dismisses the bridge team’s squat tables as an irrelevance that can be ignored, especially if he then identifies points during the intended passage where the UKC is likely to be less than the tabulated squat! The intention of this article is to try to increase the overall understanding of the squat phenomena.

Freight Ro-Ro. Draft 6,5m Speed 10 kts UKC approx 8m Photo : JCB

 

The same ship. Speed 20kts UKC 10m Photo JCB

My interest was triggered by a major feature on ship squat, somewhat alarmingly titled “Don’t Fall Victim To Ship Squat Perils” in the July 2006 issue of the NAUTILUS Telegraph, written by Dr. Barrass FNI whose mathematical calculations and tables are those used throughout the Industry and shipping fleets. In that article Dr. Barrass reproduced tables which indicated that at 10 kts speed through the water a vessel with a high block coefficient such as a tanker or bulk carrier would squat between 1 and 2 metres and that this might be doubled if another vessel was passed in shallow waters. This is clearly an incorrect figure because such vessels have been safely transiting shallow water port approaches using under keel clearances of 0.5m and 1.0 m years before any calculations had been produced to suggest that such parameters were unsafe! I had therefore been planning to write and question Dr. Barrass’ mathematics myself but Houston pilots’ representative, Louis Vest, beat me to it and the following is an extract of his letter:

… we will transit a vessel with 13.7m of draft and 0.6m of under keel clearance. A typical transit speed for such a vessel will be about 10 to 12 knots across the bay.

According to the author’s tables, we should experience about 2m of squat, but we don’t. The ship runs up the channel, the 0.6m under keel clearance doesn’t change, and we deliver the ship safely to her berth. This is not a rare event but a daily occurrence. Contrary to the author’s claim that squat increases in shallow water, squat appears to disappear in very shallow water.

The author also asserts: ‘The presence of another ship in a narrow river (passing, overtaking or simply moored) will also affect squat — so much so that squats can double in value as they pass/ cross the other vessel.’ We make our transits with two-way traffic. In no case has a change in squat been a factor in these meetings. I do believe that squat exists, but squat and ship hydrodynamics in very shallow water are a very poorly understood phenomenon. If the author would like to correspond with me in the interest of clarifying these points, I would be happy to oblige.

 

Such observations are in accordance with our experience in London and many other major ports such as Rotterdam, also safely undertake passages using low UKC parameters so one would have expected that Dr Barrass would take up the offer of dialogue with the Houston pilots in order to try to understand and resolve the anomalies between the actual and theoretical squat. However, rather than entering into a constructive dialogue, Dr Barrass chose to respond in the December 2006 Telegraph with another article titled “Ship Squat Is A Real Issue In the Real World” where he used his complex formulae to reveal how the Houston pilots totally misunderstood how ships behave in shallow waters and narrow channels and accused them of dangerously negligent navigation practices which couldn’t possible be undertaken without a grounding. Unfortunately, Dr Barrass’ diatribe reveals almost total ignorance of real ship operations with, for example, the following responses to Louis Vest’s comments regarding transit speeds of 10 – 12 kts:

To me, his (Louis) speed appears to be ‘ship speed over the ground.’ This is the speed measured when using GPS. It is not the speed that I use, namely the ‘ship speed relative to the water.’ Louis Vest has mistakenly ignored the speed of current flow. At zero current flow it is not possible in hydrodynamics to have to a ship speeding at 10 to 12 knots along a channel where B/b is about 4 and H/T is about 1.04. In the real world the local port authority would take an extremely dim view of these speeds. Furthermore, the machinery space within an oil tanker would not generate sufficient power to produce these ship speeds along this channel.

 

With respect to Louis’ comments regarding squat seeming to disappear at slow speeds Dr. Barrass reveals poor research methodology by stating:

 

This just cannot be so. It defies the laws of physics. It contradicts the laws of the Venturi effect. I have a database of 69 vessels that have gone aground due to ship squat problems. If this quote were true, then we would not have had any grounding such as the Herald of Free Enterprise in 1987, the QE2 in USA in 1992 and the Sea

Empress in 1996.

These are unfortunate examples because only the QE2 grounding was directly attributable to squat and checking on other examples of groundings listed by Dr Barrass as having been caused by squat reveals that at least two ( Tasman Spirit and Diamond Grace) grounded for reasons entirely unrelated to squat and most of the others listed cannot be directly attributed to squat. However, Dr Barrass’ arguments were reinforced by retired Venetian pilot, Sergio Battera MNI who agreed with Dr Barrass that UKC of 0.6m at speeds of 10 – 12 kts would be unsafe and could result in a grounding!

Louis Vest obviously responded and the following extracts highlight the key facts regarding the everyday navigation practices at Houston:

a. Dr Barrass provided a table showing predicted squat of 1-2m for a vessel travelling at 10 knots in a confined channel.

The Houston Ship Channel crosses Galveston Bay for over 30 miles. It is a man-made channel 530ft wide and 45ft deep in a bay whose average depth immediately

outside the channel is around 12’ft, making this a restricted channel.

b. He made the assertion that the predicted squat can double in value as one vessel passes another vessel.

The project depth of the channel is 45ft and the initial dredging was to 47ft, as

measured with tide value = 0. c.

c. Vessels making too great a speed in shallow water will ground due to squat.

We accept ships for transit to Houston with drafts of 45ft at 0 tide, 44ft at -1 tide, etc…

d. The lower the value of underkeel clearance the greater the value of predicted squat.

We transit Galveston Bay at speeds of 10-12 knots in these deep draft vessels.

We do not run aground.

We operate in a two-way traffic environment and do not ground when meeting other vessels, even similarly loaded vessels. These are not calculations or predictions as some have suggested (Capt. Battera of Venice, January Telegraph). They are easily verifiable facts. They are not exceptions or rare occurrences but everyday events in the busiest port in the United States. The fact that our daily practice runs counter to accepted theories of squat is somewhat unfortunate for the scientists who have made this their life work, but it is no less true because of it.

Dr Barrass attempts to discredit my letter in several ways. In one paragraph he insultingly suggests that I am confused about the difference between speed over ground and speed through water. I would like to assure Dr Barrass that we humble seamen, in our crude Neanderthal way, are aware that current affects the speed of a vessel. In another paragraph Dr Barrass suggests that operating vessels at 10-12 knots across the bay is unsafe and irresponsible. The morality of the Houston pilots is not the question. We stand on our safety record. The ship, in a strictly scientific sense,

cannot act irresponsibly. As an inanimate object (regardless of what personality traits she might manifest for her crew) the ship makes her transit in complete innocence of Dr Barrass’s opinion of her conduct. For Dr Barrass to suggest that crossing the

bay at 12 knots is dangerous or irresponsible ignores the substance of the subject, which is that Dr Barrass has published a table that says it can’t be done and yet it is done on a routine basis. Elsewhere in his letter Dr Barrass employs his formulae to

assert that the ship’s machinery is inadequate to propel a vessel through a confined channel at 10-12 knots. Your readers can decide that one for themselves. On one hand you have a formula with B/b = 4 and H/T=1.04 and who knows what. On the other hand there is the vessel herself in one place one moment and a mile away still going strong five minutes later. I personally think Velocity=distance/time trumps Dr

Barrass’s hydrodynamic formulae here. Dr Barrass also uses formulae (not given) to predict that a vessel with a given speed in deep water would have her speed reduced

to about one-third that speed in a confined, shallow channel given the same engine input. In practice, a vessel running very close to the bottom as we are discussing will make about 80% of the posted speed, not 33%. As the underkeel clearance increases

that percentage approaches 100%. For example, a large tanker making turns for 12 knots as indicated in her tables will make about 10 knots. The same tanker outbound in ballast will make 11-11.5 knots. A small coaster with 5-6m of draft will make very close to her indicated speed. Yes, I am allowing for current (see above). In order to burst Dr Barrass’s hopes that a change in water density from fresh water to salt water might explain our miraculous escape from the laws of physics as he interprets them, I would like to point out that Houston is a major oil port and the United States is an oil

importing nation. Most, but not all, of our deep draft transits begin in the salty water of the Gulf of Mexico and terminate in the fresh water of Buffalo Bayou. The claim is made that we would not have had groundings such as the Herald of Free Enterprise, the QE2, and the Sea Empress if not for the effect of squat. According to the official

report, the Sea Empress grounded on rocks due to the pilot’s failure to adequately allow for the set of current across the channel. The Herald was notoriously lost due

to the failure of her crew to secure her bow doors properly. Neither accident was remotely related to squat. The QE2 struck a rock jutting up from much deeper

water all around. While this accident involved squat it is not related to my

observations about squat in a confined channel with a continuous minimum underkeel

clearance.

Now mariners from the ends of the earth (Houston) are telling scientists that their predictions of ship behaviour do not match real world observations. I suggest we

recognise that our knowledge of the hydrodynamics of large vessels in very shallow water is indeed poorly understood. This represents an opportunity to advance our understanding of the world if properly taken.

This situation is obviously unsatisfactory. On the one hand we have pilots safely bringing ships in and out of port with minimum UKC parameters and on the other hand there are the scientific “experts” who have produced tables to prove that what we are doing is impossible. There is no other industry that would not only accept the mathematics without question but also create a safety policy around such poor scientific analysis. I have never met Dr. Barrass and I am sure that he is a very eminent mathematician but I would have thought that rather than publicly accusing the Houston pilots of incompetence and revealing a somewhat alarming ignorance of real ship behaviour, a curious scientist would have undertaken further research to explain the anomalies between the mathematical predictions and real time practical experience. To be fair to Dr Barrass his calculations are reasonably representative of others working on this phenomena and I will at least give him credit for sticking his head above the parapet and stimulating some emotional debate!

So, what is squat?

The Permanent International Association of Navigation Congresses (PIANC) (see Autumn 2007 issue) is now the main forum for squat related issues. This is a positive development because the UKMPA attend the PIANC sessions so pilots’ now have direct input into the discussions. The PIANC papers provide the following definitions of squat. Squat is the reduction in UKC between a vessel at-rest and underway due to the increased flow of water past the moving body. The forward motion of the ship pushes water ahead of it that must return around the sides and under the keel. This water motion induces a relative velocity between the ship and the surrounding water that causes a water level depression in which the ship sinks. The velocity field produces a hydrodynamic pressure change along the ship that is similar to the Bernoulli effect in that kinetic and potential energy must be in balance (Newman 1977). This phenomenon produces a downward vertical force (sinkage, positive downward) and a moment about the transverse axis (trim, positive bow up) that can result in different values at the bow and stern (Figure 1).

Most of the time squat at the bow S_b represents the maximum value, especially for full-form ships such as supertankers. In very narrow channels or canals and for high-speed (fine-form) ships such as Passenger Liners and Containerships, the maximum squat can occur at the stern S_s. The initial trim of the ship also influences the location of the maximum squat. The ship will always experience maximum squat in the same direction as the static trim (Barrass 1995). If trimmed by the bow (stern), maximum squat will occur at the bow (stern). It is the classical “Venturi Effect” as streamlines will move faster under the smaller cross-sectional area at the bow (stern) resulting in lower pressure (i.e., more suction) and increased squat. It is not possible to compensate for increased squat at one end by trimming at the other end.

Factors Governing Squat

Prediction of ship squat depends on ship characteristics and channel configurations. The main ship parameters include ship’s draft, hull block coefficient, and speed through the water. The main channel considerations are proximity of the channel sides and bottom Channel bends and proximity to banks tend to increase squat and muddy bottoms to decrease it. The presence of another ship (passing or moored) can also increase squat.

The most important ship parameter is its speed through the water and generally squat varies as the square of the speed so doubling the speed quadruples the squat.

The most important channel parameter is the water depth and can be ignored if the depth is twice the draft or more.

How is squat calculated?

Consider yourselves lucky that I am not going to reproduce the mathematical formulae used to calculate squat because they are pretty baffling to a simple seaman such as me but for those of you who would like to see them they can be found via the web links I have placed at the end of this article.

The calculations are made for three different shallow water conditions shown above and several specialists have produced mathematical formulae to calculate squat but Dr Barrass’ formula are the most well known and widely used.

However, they all use the same basic concept so produce similar predictions but as highlighted by the exchange between dr. Barrass and the Houston pilots there are seemingly serious anomalies between the predicted squat and the actual squat experienced.

The following are examples of calculated squat from the ten main study groups for a selection of typical vessels illustrating the range between the results. They are for bow squat in an unrestricted channel with an initial UKC of about 2.5m and speed through the water of 11 kts. The name of the research group is in brackets.

250,000 DWT Tanker

Largest squat (Milward) 1,25m

Least squat ( Eryuzlu) 0.50m

(Barrass) 0.80m

65,000 DWT Tanker

Largest squat ( Milward) 1.10m

Least squat (Romisch) 0.43m

(Barrass) 0.90m

Panamax Container Ship

Largest squat (Milward) 0.75m

Least squat (Romisch) 0.20m

(Barrass) 0.30m

As can be seen there is a considerable difference between the various researchers so the allegation by the Houston pilots that squat is a “very poorly understood phenomenon” is fully justified.

Resolving the anomalies.

The existing calculations are nearly all based upon theory or laboratory testing methodology and PIANC has recognized this and produced the following recommendation regarding squat :

PIANC recommends model tests for specific ship and channel conditions, especially if the conditions are new or novel. Many laboratory-based formulas are from captive towed tests that introduce unintended moments that can cause unrealistic trim of the towed models. The current thinking is to use free-floating, remote controlled models for physical model tests. Finally, full scale measurements are always a good check of design stage predictions.

Until very recently the complexities of measuring the real time squat of ships under way were too great to enable any meaningful results. Fortunately the advent of DGPS and other technologies such as tide rate / height monitoring has enabled real time squat measurements to be made by placing specialist equipment on board which can generally produce accuracies to +/-10cm and the latest equipment is capable of accuracies to +/- 1 cm. Much of the pioneering work has been undertaken by the Australian company OMC who have developed and registered as a trade mark the concept of Dynamic Under Keel Clearance (DUKC) ® to permit maximum loadings of bulk carriers. The DUKC® concept doesn’t just allow for squat but also has to factor in the large swells that are a frequent problem in Australia and new Zealand. With dedicated precise instrumentation fitted on board several bulk carriers and precise swell meters and tide gauges sited at critical points along the intended route, data is fed into computers at the loading terminal and these ships are thus loaded to the absolute maximum for the existing conditions. I understand that this loading method has resulted in additional cargo liftings of around 1,500,000 tonnes per annum at just the Hay Point terminal in Queensland alone.

Other real time measurements have been undertaken by the US Army Engineer Research Centre (USACE) and on the Elbe both live trials and specialist model tank tests have been undertaken by the Federal Waterways Engineering and Research institute (BAW).

 

Results from Real observations and specialist “free” model tank tests.

 

Although the number of real ship trials is still fairly low there is now real data beginning to emerge that confirms that the existing predictions are over pessimistic which of course comes as no surprise to pilots! Full details of many trials can be obtained via the links at the end of this article but the following are a small sample of results from trials undertaken by the above groups. Please note that these are very basic interpretations from detailed graphs.

.

 

BAWTank test:

Panamax Containership: Draft 12.8m, Initial UKC 5.7m

Speed Squat

10kts 0.3m

12kts 0.4m

14kts 0.7m

 

Bulk Carrier (350m loa) Draft 14.5m Initial UKC 4.0m

Speed Squat

8kts 0.5m

10kts 0.7m

12kts 1.0m

BAW Live trials.

Data was collected from 9 transits of large container ships on the Elbe. The results reveal an interesting difference between wide and normal transom ships with the wide transom displaying far less squat.

Speed least observed squat largest observed squat

10 kts 0.2m 0.5m

12 kts 0.5m 0.9m

15 kts 0.6m 1.7m

USACE observations

In 1999 the USACE undertook live trials in Charleston. The following is a sample of the results which compares the observed and predicted bow squat.

Predicted squat

Ship Observed Squat Huuska Barrass Romisch

PX Container 0.99m 1.86m 1.99m 0.99

190m Bulk Carrier* 0.53m 1.03m 0.94m 0.66

* Restricted Channel.

OMC Observations

OMC have been undertaking real time trials using very precise measurements. Recently they have completed real time trials for the port of Port Marsden in New Zealand. The results are very detailed because the DUKC® concept also factors in swell and other environmental conditions but as an example, measurements on a 100,000 DWT tanker with a speed through the water of 9 kts in a restricted channel gave an observed squat of about 0.4m.

My own observations

Totally unscientific but with nearly 20 years of piloting ships from small coasters to VLCC’s in restricted channels my personal observations are:

• At speeds of less than 9kts through the water squat is negligible.
• Squat is to be taken seriously if the vessel is passing rapidly from deep to shallow water.
• If a vessel is already in shallow water then the ship and environment will warn you (engine vibration and breaking quarter wave) that you are going too fast before a grounding occurs due to squat
• Modern pitch control propeller systems have overload protection that will prevent excessive speed in shallow waters
• It is important to discuss the UKC with the Captain! He will have been in and out of many ports and will normally have a good understanding of how his ship behaves in shallow waters especially if he has transited the Houston ship canal and played “Texas Chicken” (www.texnews.com/1998/2002/texas/texas_Pilots_Se822.html) after which very little will perturb them!!

Conclusions

Squat is an extremely complex subject and is dependent on many factors including mud*. Practical trials generally reveal the traditional tables to be over estimating squat which of course provides a safety margin. However if they are so inaccurate that they are unrepresentative of reality then they are an alarmist waste of time. What I find difficult to comprehend is that despite many real time observations indicating anomalies in the tables, these tables are still being provided for every ship and potentially introducing conflict in the Master / Pilot exchange. Fortunately most Masters accept that the port’s own established guidelines, applied by properly trained pilots will result in safe transits but the overall situation is unacceptable.

http://chl.erdc.usace.army.mil/library/publications/chetn/pdf/chetn-ix-14.pdf

http://www.omc-international.com/

Dr. Barrass: www.ship-squat.com

*Next quarter I will be examining the linked concept of mud navigation and navigable bottom!

Feedback from Louis Vest:

I’m glad to finally get some support from other pilots. Interestingly, Dr. Barass; after trying to discredit me in the magazine with his second article, has never contacted me about my offer to observe squat (or lack thereof) in Houston. Nor has any other hydrodynamacist. Regarding the reliability of models. Another of my disagreements with the professorial experts has been the validity of observations/tests/predictions based on data obtained from test tank models or mathematical models. Validation of these models is important because ship designs, rudder designs, etc are increasingly made and tested on computers. In some cases portions of sea trials are allowed to be simulated or performed in a test tank. In one of the numerous attempts to verify squat calculations that you described above, the USCOE funded an effort here in Houston. The goal was to install highly accurate GPS units (+/- 2cm)on the bow, stern and beam ends of vessels making Houston transits. The recorded data was to be made available online for investigators to use. Last time I checked no one had done any work on the data. More importantly, there were several ships calling in Houston at the time for which mathmatical models were available; the same models used for generating the graphics of simulators for example. The researchers made the extra effort to record data from these vessels and in at least one instance two vessels met in the channel, both of which were equiped for data recording and both of which had known mathmatical models. When the ships met bridge cameras were used to record the exact timing and sequence of pilot orders during the meeting. (For those who aren’t familiar, a meeting in Houston involves two ships meeting head on until about .5 mile and then swinging right to pass within 60-70 meters of each other in a 160m channel.) After the study, two of the researchers tried to recreate the maneuver on a simulator using the ship models, the recorded sequence of commands, and the recreated channel outlines. The results did not conform to the actual meeting situation. The simulator ships could not recreate the maneuver. The reported the results were significantly different on the simulator from real life. I tried to convince them that a negative result was important and worth an article in one of the leading scientific journels, but they didn’t seem to think so. So. How accurate is simulator training or simulator ship design? It’s not just about squat. Louis Vest.