Read 24.pdf text version

Trans IMarE, Vol 109, Part 3, pp 257­269

Ballast system design for flow-through exchange of ballast water

G Armstrong, BSc, PhD, CEng, FIMarE

Three Quays Marine Services, London

Ballast water discharge has been identified as a potential risk to health and a major cause of ecological disturbance due to the discharge of pathogens and non-indigenous species of bacteria, plant and animal life at the end of long ballast voyages. The Marine Environment Protection Committee (MEPC) of IMO is considering these hazards, and a working group is investigating means by which the risks can be minimised. To date, international guidelines have been promulgated as IMO Resolution A.774(18), annexed to M Notice 1533, and a new MARPOL Annex is envisaged, containing regulations supplemented by a revised set of guidelines. A requirement to exchange ballast water at sea must take into account the implications for stability, longitudinal strength, and the risk of significant structural loads due to sloshing in partially filled tanks. The paper presents the results of case studies applied to bulk carriers and containerships to examine the design of ballast systems which can provide continuous sea-to-sea circulation of ballast water while the ballast tanks remain full. System design details and additional costs are presented, and the application of continuous circulation to typical voyages is considered.


There is now a large body of evidence which shows that species of bacteria, plants and other organisms can survive voyages of several weeks duration in ships' ballast water. Subsequent discharge of such ballast or sediment can result in the establishment of unwanted non-native species which can upset the local ecological balance. It is also possible for diseases to be introduced into Port State waters by discharge of infected ballast water. This potential for ballast water discharges to cause harm was recognised in Resolution A774 (18) of IMO, and guidelines were issued for the prevention of the introduction of unwanted aquatic organisms in ships' ballast water and sediment discharge. These guidelines are promulgated in the Annex to Merchant Shipping Notice M1533.


One of the options being considered at the IMO is a new Annex to Marpol 73/78 which will contain regulations and a revised set of guidelines. In their present state of development the regulations will require a Ballast Water Management Plan to be drawn up, which is to include recommendations based on the revised guidelines and recommendations from the Design and Equipment Subcommittee relating to safety issues for the exchange of ballast water. Ships will also

Author's biography

be required to maintain records of ballast uptake and discharge, including date and time, geographical location, quantities, salinity and specific gravity. The condition of any clean ballast is to be confirmed by a certificate issued by the Port State Authority where the ballast has been loaded, if acceptable to the receiving Port State Authority. Operationally, uptake treatment and discharge of ballast water and sediments are to be in accordance with the guidelines. If a ship cannot comply due to weather, sea conditions or any other condition which in the Master's opinion would threaten the ship's safety, a report is to be made to the Port State Authority so that alternative measures can be taken. Last November, the United States passed the National Invasive Species Act of 1996, which requires that vessels entering US waters from outside the 200 mile Exclusive Economic Zone (EEZ) must exchange their ballast water before entering the EEZ. The procedure is initially voluntary, but would become mandatory after two years if compliance was deemed to be insufficient. In addition to the present international guidelines contained in M 1533, MSA has issued M Notices 1532 and 1662 which reproduce the guidelines promulgated by Australia, New Zealand and Israel concerning the control of ballast water discharges destined for their ports. The requirements of M 1532 in respect of Australia and New Zealand are similar to the international guidelines in M 1533, and Israel requires that a ship must exchange any ballast water not taken onboard in open ocean. Ships visiting Eilat must exchange outside the Red Sea and those visiting Mediterranean ports must exchange in the Atlantic.

G Armstrong graduated from the University of Durham in 1973 with a degree and PhD in Engineering Science. Between 1973 and 1977 he held a Short Service Commission as an Engineer in the Royal Navy. From 1977 to 1989 he lectured in marine engineering at the University of Newcastle upon Tyne, undertaking research in the performance of residual fuels in medium-speed engines. He joined Swan Hunter Shipbuilders as Head of Marine Engineering Basic Design in 1989 and joined Three Quays as Manager, Engineering, in 1994.

Paper read on 25.03.1997


G Armstrong

Since 1994 the IMO Ballast Water Working Group has been reviewing the international guidelines, and the present revision discusses treatment options in greater detail than those in M 1533. Several of the following options are straightforward statements of good practice but in many circumstances the choices available to an operator may well be very restricted, depending on technical and scientific advances.

Ballast water age

Absence of light and the higher iron content of water in ballast tanks provide a hostile environment for many organisms, and the guidelines suggest that water which has been in a ballast tank for longer than 100 days represents a minimal risk although toxic dynoflagellates can survive for several years. Further research is required, but retention of ballast water for periods in excess of three months is unlikely to be a realistic option for tankers and bulk carriers.


Sediment disposal

This represents an obvious precaution and highlights such considerations as the avoidance of shallow water (including propeller wash in shallow water), the prevailing turbidity in relation to ebbing or flowing tides, avoidance of dredging operations, and avoidance of areas of known outbreak of disease or phytoplankton blooms. It is acknowledged in the guidelines that there will be circumstances in which an operator has little choice concerning the quality of ballast taken onboard. The guidelines recognise that many aquatic organisms are present in sediment and that all sources of sediment retention, such as anchor cables, chain lockers and suction wells, should be routinely cleaned. Design of internal structure should also be such that sediment retention is minimised.

Reception facilities

The guidelines state that where shore reception facilities exist, discharge of ships' ballast water into such facilities may provide an acceptable means of control. To say the least, this is unlikely to be a practical option as shore reception facilities for other waste materials are generally considered to be unsatisfactory.

Certificate of cleanliness

Ships bound for ports which are considering the implications of the new Marpol Annex may be able to avoid at-sea exchange of ballast water by obtaining a laboratory analysis of the ballast water, and providing a certificate if the water is found to be free of unwanted aquatic organisms or pathogens considered harmful by the receiving state. Realistically this cannot be an effective method of active risk minimisation and would be more suited to a liner trade,asprior knowledge of the acceptability of a certificate to the receiving state would generally be required. It also presupposes that the ballast water in the area where it is taken up is acceptable to the receiving state, and is likely to remain so for a period of time.

Ballast water exchange

Ballast water exchange is the most effective practical method of minimising the risk of transfer of unwanted organisms. Compared with coastal and estuarine waters, deep ocean water contains few organisms, and those that do exist are unlikely to survive a transfer to a coastal or fresh water environment. The guidelines recommend an exchange in water depths of at least 2000m. However, whereas ballast exchange would be most effective if the tanks were completely emptied and refilled, this may not be possible for reasons of stability, longitudinal stresses, or the risk of damage due to sloshing action in partially filled tanks. In such cases, the guidelines suggest that the ballast tanks can be overflowed so that at least three times the tank volume is pumped through the tank. Encouragingly, this figure of three times the tank volume is not merely an estimate, but is supported by experimental results1.

Non-release of ballast water

Obviously the most effective means of preventing the introduction of unwanted organisms is to avoid the discharge of ballast water. In containership operations this may be achievable in ports where containers are both loaded and discharged, but it is not an option for bulk carriers and tankers.

Changes of salinity and changes of temperature

The ability of aquatic organisms and certain pathogens to survive transportation in ballast water may be reduced if there are significant changes in ambient conditions. It is acknowledged that further research is necessary to establish the efficacy of salt water/fresh water changes, and the temperature differences required for effective treatment. At present, any beneficial effect can only be regarded as fortuitous as it is outside the operator's control.



The guidelines note that there is a clear need to research and develop additional measures for ballast water treatment, particularly as new information on organisms and pathogens is becoming available. Suggested areas are:

Trans IMarE, Vol 109, Part 3, pp 257­269

1. treatment by chemicals, biocides and UV radiation; 2. heat treatment; 3. oxygen deprivation control; 4. tank coatings; 5. filtration. As a contribution to the discussion at the IMO, a comprehensive review of the options for ballast water treatment has been made by Lloyd's Register, at the request of the MSA2. The treatment options are divided into three categories of mechanical, physical and chemical options, and the advantages and disadvantages of the more likely options are summarised as follows.

Based on this, tests have been undertaken to examine the effects of lower temperatures for extended periods of time. Temperatures as low as 35°C and treatment times of between 30 min and 5h appear to be effective, and further trials on heat treatment are in progress. The potential application of ultrasonics for ballast water treatment has not received detailed consideration, and is likely to be, at best, only partially effective. There are disadvantages with respect to health and safety due to noise from some transducers, and due to prolonged exposure to ultrasound.

Chemical treatment

The use of chlorine as a disinfectant and the use of electrochlorination as a means of preventing marine growth in seawater systems is well known, and the principle could be extended to the treatment of ballast water. For several reasons, however, electrochlorination, or the addition of chlorine by other means, is not considered a preferred option for ballast water treatment. Firstly, the effectiveness of chlorine as a disinfectant is difficult to predict in all circumstances. The exact mechanism by which it kills aquatic organisms is not known and chlorine will react with ammonia, organic matter, iron and manganese. The presence of such material substantially increases the dosage rate required to achieve a given level of free chlorine available for disinfection. Effectiveness is also influenced by temperature, contact time and pH level. Apart from the corrosive effect of free chlorine, there are environmental concerns about the discharge of chlorinated ballast water, and it is known that the reaction of chlorine with some organic compounds produces carcinogens. Also, the release of hydrogen by electrochlorination on the scale required for ballast water treatment is a substantial safety disadvantage. Electrolytically generated copper and silver ions would be better than chlorination. Environmental implications would need further consideration in relation to the concentrations required, but health and safety issues onboard ship appear to be minimal, and the effectiveness of copper and silver is superior to chlorination. Some organisms can enhance their tolerance to high concentrations of copper and silver but the technique of recirculating ballast water through an electrolytic cell appears to have potential. Overall, the Lloyd's/MSA study concluded that there was no obvious solution to the question of ballast water treatment, but that some techniques may be useful either in combination or in support of other procedures discussed in the guidelines. Although all of the treatment proposals raise environmental acceptability questions, UV radiation and heat treatment are possible contenders, and of the chemical methods only the electrolytic generation of copper/silver ions appears likely to be acceptable. Straining or filtration would be required before any treatment could be effective.

Mechanical processes

Filtration is considered to be potentially helpful in removing large particles such as small seaweeds, but removal of smaller organisms would be poor. This method is attractive as the residues would be disposed of in the area of ballasting, thus avoiding the transfer of non-native species, but the capital costs (£500 000­£750 000) are likely to be high. Cyclonic separation was also investigated, but appeared to be limited in effectiveness with regard to large particles, due to the similarity in specific gravity of many organisms to that of seawater.

Physical treatment

Ultra violet (UV) radiation emerged from the study as a candidate for further research. Its effect varies with the type of organisms, and some are highly resistant to UV radiation. However, in combination with the removal by filtration of larger particles, it is likely to be acceptable on environmental grounds as there are no toxic side effects, and there are no adverse effects on pipework, pumps or coatings. Ballast water heating is considered potentially attractive as a means of at least partial sterilisation. Exposure to temperatures of 36­38°C for between 2 and 6h will kill zebra mussels, and in shipboard experiments exposure to temperatures above 40°C for approximately 8 min was shown to be lethal to all identified organisms, but not cholera. The applicability of ballast water heating would depend on the availability of sufficient heat in order to treat the water within the duration of a voyage. Thermal stresses and the possible implications for tank coatings would also need to be addressed. There may also be environmental objection to the discharge of warm water. Taylor3 considers that the best available treatment option is ballast water exchange, but the effectiveness can be further improved if other treatment options are combined with it. Heat treatment is a strong candidate, and an analysis based on the 141 000 dwt bulk carrier Iron Whyalla showed that heat could be utilised by flushing water at 45°C through the ballast tanks in sequence, allowing the excess to overflow. There was, however, insufficient heat available to reach the preferred temperature of 40°C: for ocean temperatures of 25­30°C equilibrium temperatures of 35­38°C may be possible after flushing each tank for 48h.


All of the options for treatment refer to sterilisation of the ballast water so that it can be safely discharged into the


G Armstrong

receiving area. Much the better solution would be to exchange contaminated ballast water at sea for relatively inert, organism free, deep ocean (>2000m) seawater. Exchange by means of emptying and refilling tanks would be preferred but the guidelines recognise that this may not always be achievable, due to considerations of stability, longitudinal stress limits, and the possibility of damage due to sloshing in partially filled tanks. Rigby and Hallegraeff1 examined the effects of ballast water exchange by emptying and refilling tanks on the shear force and bending moment of Iron Whyalla. Under conditions in which one set of tanks (port and starboard double bottom and topside) is empty, with all others full, it is shown, for example, that the still water bending moment can be as high as 163% of the maximum allowable. Peak stresses in rough weather would of course be greater. Additionally, as tanks further aft are emptied, propeller immersion is reduced to a degree which is unacceptable. This leads to consideration of the alternative of flushing the tanks with at least three times the tank volume, following the experimental results1 on Iron Whyalla. One of the series of trials conducted on Iron Whyalla was to examine the process of ballast water exchange. Methylene blue dye was added directly to the tank through three sampling tubes and a period of 20h allowed for dispersion. Samples were taken from the three points in the tank: one inboard in the double bottom adjacent to the pipe tunnel (although the longitudinal position of this sampling point relative to the suction/discharge is not stated); a second in the double bottom outboard; and a third in the topside tank. For the purpose of the trials the tanks overflowed through a manhole access on deck. The theoretical concentration in the tank during flushing was calculated assuming perfectly mixed conditions. No modifications were made to ballast tank internal pipework to improve water distribution, and a comparison between the theoretical concentration and the concentration measured at the inboard point in the double bottom after the exchange of one double bottom tank volume (ie one residence time), showed a substantial `short circuit' component (where less than ideal mixing takes place between the original ballast water and the incoming water). However, at the overflow the measured concentration was close to that predicted by the theory, assuming perfect mixing, after one residence time. However, without distribution pipework in the double bottom to promote mixing or details of the position of the sample tube in relation to the filling point, it is possible that there remained unflushed areas in the double bottom, although flushing of other areas of the tank appears to have been thorough. Assuming perfectly mixed conditions, the efficiency of the exchange was as shown in Table I. Table I shows that after an exchange of three tank volumes, 95% of the original ballast water has been exchanged and this is the justification behind the figure used in the IMO guidelines.

Table I

Efficiency of exchange due to flushing

Proportion of original ballast exchanged (%) 39.3 63.2 86.5 95.0 98.2

No of tank volumes exchanged 0.5 1.0 2.0 3.0 4.0

Table II Ormond ballast tank capacities

Tank F peak No 1 P No 1 S No 2 P No 2 S No 3 P No 3 S No 4 P No 4 S No 5 P No 5 S Aft peak Total Capacity (m3) 5376 2444 2444 6430 6430 6478 6478 6465 6465 4695 4693 1170 59 568

190 000 dwt P&O bulk carrier Ormond, which trades between Europe, Brazil, Japan and Australia. The voyages from Europe to Brazil (14 days) and from Japan to Australia (11­12 days) are ballast voyages. Ballast tank capacities are given in Table II. In addition to the ballast tank capacity in Table II, No 6 cargo hold, of capacity 21 811m3, is always filled when the ship is in ballast, giving a total ballast capacity of 81 379 m3.

Exchange duration

There are two ballast pumps, each with a capacity of 2500m3/h, and hence, with one pump running, the duration of flushing to achieve three times the tank volume is: 3 x 81383.3 h = 97 h 2500 (1)

or 4.07 days. This can be achieved within the 14 day Europe to Brazil and 11­12 day Japan to Australia ballast voyages.

System design

The port and starboard double bottom and upper hopper ballast tanks are shown schematically in Fig 1. The upper hopper tanks are permanently connected to the double bottom tanks by trunks at each end. Ormond is not fitted with a duct keel: the ballast main runs through the ballast tanks as shown. For continuous flushing purposes, and to promote distribution in the double bottom, it is proposed to fit an internal arrangement of pipework in order to provide a point of discharge between each solid floor. The diameter of this pipework has been calculated so that approximately


A detailed design of a ballast system to permit continuous flow exchange of ballast water was investigated for the


Trans IMarE, Vol 109, Part 3, pp 257­269

Fig 1 Bulk carrier ballast system modifications

equal flow rates are achieved at each discharge point. Such an arrangement merely minimises the risk of areas of stagnation in the tanks and the risk of sediment deposition. There is no intention to scour the tank surfaces with any impingement equivalent to crude oil washing. Bulk carriers which ballast in deep water ports are less likely to take on large quantities of harbour sediment, and after 10 years service the tanks of Ormond were seen to be remarkably clear. A similar observation was made with respect to Iron Whyalla1, although P&O report that Ormond is not necessarily typical and ballast tank sediment is a significant problem in some ports. From each distribution point flow is outboard and up into the hopper tanks, from where it is discharged overboard above the ballast water line through dedicated overboard connections. The height of the entry point to the discharge pipes within the tank is calculated so that the tank remains full with the flowrate of one ballast pump. Water is not discharged onto deck through the tank vent pipes. Lloyd's

Register have raised no objection in principle to the proposals, and confirmed that the overboard discharge valves would not be required to be non-return. The Rule-sized air pipes remain unchanged. Discharge overboard and the supply to the internal distribution pipework in the double bottoms are controlled by remotely operated valves. The arrangements for No 6 cargo hold are different as no internal pipework can be fitted in a cargo hold, but there is, of course, no solid flooring to impede mixing as there is in a double bottom tank. This is shown in Fig 2. The water level is to be maintained just under the hatch, but the presence of steel structure inhibits access for pipework directly to the hatch coaming. Instead, the connection to the hold is taken into the adjacent hopper tank, with the height of the right angle bend in the tank designed so that the water level in the cargo tank is maintained under the flow from one ballast pump. The highest part of the bend is fitted with a flushing connection in case of ingress of cargo.


G Armstrong

Fig 2

Cargo hold discharge

Fig 3 Fore peak tank

The fore peak tank at 5376m3 contains a large proportion of the ballast total, and it is therefore proposed to fit internal pipework to facilitate mixing (Fig 3). Discharge pipework port and starboard is fitted in a similar arrangement to the ballast tanks. In the after peak tank, which contains only 1170m3, the supply and discharge can be well spaced longitudinally and


hence internal pipework to promote mixing would not be required (Fig 4).

Ballast system modification costs

For ballast exchange on this vessel a glass-reinforced plastic (GRP) piping distribution system would be proposed in

Trans IMarE, Vol 109, Part 3, pp 257­269

Fig 4 Aft peak tank

Fig 5 Suezmax tanker ballast system

each tank, supplied through a 350 mm bore valve. Overboard discharges from each tank are through two 350 mm bore ship side valves with galvanised steel piping. All valves are remotely operated from a central control panel. Pipes are sized on the basis of the flowrate from one ballast pump. The ballast system modification costs are as follows: 1. GRP distribution pipework in FP tank and ballast tanks, including installation £260 000 2. Overboard discharge pipework, including installation £154 000

3. 23 shipside flanged butterfly valves 4. 11 tank distribution wafer butterfly valves Total, including valve actuation and remote operation

£80 500 £33 500 £528 000


Arrangements for continuous flushing in tankers would be similar to those for a bulk carrier. Figure 5 shows the layout


G Armstrong

Fig 6 VLCC ballast system

for two cargo tanks athwartships, with single double bottom and side ballast tanks. An internal distribution main with multiple discharge points would be provided to promote mixing. Arrangements can also be made to incorporate exchange of ballast in the centre tank of a VLCC (Fig 6) if it is used as a gale ballast tank.

harbour of discharge, and clean ballast quantity and tanks used. Nevertheless, the principle of continuous flushing can equally be applied to a containership.

Exchange duration

Ballast tank capacity 16 612.9m3 Ballast pump flow rate 500m3/h With one pump running the duration of flushing to achieve three times the tank volume is:

3 x 16 612.9 500 x 24


The P&O Jervis Bay class are 4000 teu capacity containerships with an overall length of 292.2m. Ballast tanks on each side of the hull provide a double skin configuration and a total capacity of 16 613 m3. Unlike tankers and bulk carriers, which are either fully laden or fully ballasted, a containership will most often only partially discharge a cargo, or loading and discharge may take place simultaneously, with very little net change of ballast condition. The Jervis Bay class trade between Europe and SE Asia and arrive fully loaded in Southampton, where a partial discharge takes place without intake of ballast. At Rotterdam and Hamburg loading and unloading take place simultaneously and normally no ballast needs to be taken onboard. Rotterdam is also a bunkering port, and with careful management it is not normally necessary to take water ballast onboard. Simultaneous loading and discharge normally takes place in Middle East and Far Eastern ports, with fuel bunkering in Singapore. Ballast water exchange problems are therefore much less likely to arise in containerships, and it has in any case been P&O policy since 1989 to maintain records of ballast water uptake and discharge, which include the name of the port where ballast was taken, quantity and tanks used, location of



or 4.15 days.

System design

Distribution mains would be fitted in the double bottom, centre and wing tanks, and in the upper wing tank (Fig 7). Discharge pipework from the wing double bottom tank would be taken through the side bunker tanks and upper ballast tanks, using heavy gauge pipe to the overboard discharges above the loaded water line. The discharge from the centre double bottom tank could be taken between adjacent container guides to the overboard discharge valve.

Ballast system modification costs

System costs are based on a GRP piping distribution system, with each tank supplied through a 200 mm bore valve. Overboard discharges from each tank are through two 150 mm bore ship side valves with galvanised steel piping. All

Trans IMarE, Vol 109, Part 3, pp 257­269

Fig 7

Container vessel

valves are remotely operated from a central control panel and pipes are sized on the basis of the flowrate from one ballast pump. The ballast system modification costs are as follows: 1. GRP distribution pipework in FP tank and ballast tanks, including installation £165 700 2. Overboard discharge pipework, including installation £73 800 3. Tank distribution valves 4. Overboard discharge valves Total, including actuators and remote operation £29 000 £58 000 £326 500

Cruise ships

It was first thought that discharge of contaminated ballast from cruise ships would not present a significant problem, as transport of non-indigenous organisms would not arise in the case of a cruise vessel which remains in a localised cruising area, and a transoceanic cruise ship would take clean mid-ocean ballast onboard as fuel was consumed. However, the new US legislation may require ballast water to be exchanged before entering the US Exclusive Economic Zone. A typical cruise ship has a large number of small ballast tanks distributed throughout the ship. Whereas emptying and refilling the tanks would be the preferred option this may not be possible without compromising stability.


G Armstrong

Fig 8 Grand Princess ballast system

Exchange duration

As an example, Fig 8 shows the ballast tank layout of the P&O cruise ship Grand Princess (242m LBP) which has 29 ballast tanks with a total capacity of 4345m3. The largest tank is the fore peak tank with a capacity of 773m3; the smallest are No 25 and No 22 double bottom tanks with 41.4 m3 each. Using one ballast pump of 250 m3/h capacity, tank flushing of three times the tank volume would require 2.17 days.

1. Pipework costs, including installation 2. Valves Total, including actuators and remote operation

£118 900 £51 500 £170 400


There is general agreement that ballast exchange is the most effective means of minimising the spread of unwanted organisms. Indeed, current guidelines and legislation in its present form specify ballast exchange or quote it as a yardstick by which other methods are to be judged. Ballast exchange by emptying and refilling is, however, simply not safe for reasons of stability and/or structural strength. There is no obvious treatment option presently available and hence flow-through exchange becomes the most viable strategy, although the theoretical figure of 95% exchange assumes perfect mixing. In order to improve on this figure, flow-through exchange could be combined with a heat treatment process or electrolytic generation of copper and silver ions, provided these are environmentally acceptable. The modifications to provide flow-through exchange in bulk carriers, tankers and containerships are relatively straightforward, although over 50% of the costs are incurred in the distribution pipework within the tank. However, the

System design

Flushing directly overboard from each tank would result in an unacceptably large number of hull penetrations, and it is therefore proposed that separate ballast filling and emptying systems would be required in place of the common suction/discharge system. Inlet and outlet connections would be positioned at opposite ends of the tanks.

Ballast system modification costs

System costs are based on the provision of a duplicate ballast piping system in galvanised steel. Approximately 51% of the system is 150 mm bore, 42% is 200 mm bore and the remainder of 100 mm bore. All valves are remotely operated from a central control panel. No internal tank distribution pipework is included. The ballast system modification costs are as follows:


Trans IMarE, Vol 109, Part 3, pp 257­269

justification of the figure of 95% effectiveness of flow-through exchange with three times the tank volume flushing depends on ideal mixing, and the Iron Whyalla experiments showed that flushing in the double bottoms may not have been adequate, in which case internal distribution pipework would be required. In passenger ships, with a greater number of small ballast tanks, the requirement for internal pipework would depend on the individual tank configuration, but would form a lower proportion of the overall costs.

are also due to colleagues in Three Quays for their assistance, and to colleagues within P&O Companies for time spent in discussion, and for the provision of technical information.


1. G Rigby and G Hallegraeff, `The transfer and control of harmful marine organisms in shipping ballast water: behaviour of marine plankton and ballast water exchange trials on the MV Iron Whyalla', J Marine Env Eng, Vol 1, pp 91­110. 2. Disinfection of Ballast Water, a Review of Potential Options, Lloyd's Register (July 1995). 3. A H Taylor, `Design considerations for ballast water control and treatment', Paper 10, pp 23­29, IMAS 96: Shipping and the Environment: Is Compromise Inevitable?, Conference Proceedings Part I, IMarE Conference, Volume 108, 5, The Institute of Marine Engineers (1996).


The author would like to thank the Directors of Three Quays Marine Services for permission to publish this paper. Thanks


G Armstrong


E Brookes (The Chamber of Shipping) The author has given a good summary of a very difficult topic with severe safety implications. Has he considered the impact of running the pumps for the lengths of time required, the cost (in terms of fuel and maintenance) and the environmental effect of additional pollutant emissions required to generate the necessary power. Dr G Armstrong (Three Quays Marine Services, London) There will indeed be implications arising from running the ballast pumps for the required period of time. In the case of Ormond, the fuel cost for a flow-through exchange of all tanks is approximately US$470 (equivalent to 0.58 cents per tonne) but the overall power increase as a proportion of the total propulsion and auxiliary power is less than 2%. The increase in emissions is therefore small and will, of course, take place well away from land. R Lankester (Friends of the Earth International, Oceans Division) There are many voyages undertaken between discrete marine ecosystems where ballast water discharges pose an environmental threat and where the oceanic conditions necessary for exchange at sea do not exist. For example, coastal voyages, say from New York to San Francisco are coastal and if via the Panama Canal, but traverse two of the world's great oceans, the Atlantic and the Pacific. Also the voyage from the Black Sea to the Baltic via rivers and canals requires ballast water to be taken on to allow vessels to pass under bridges. The risk of translocating the comb jelly fish (Mnemiopsis) is high and, again, eradication by the flowthrough method impossible. It has been made absolutely clear by shipowners and operators that they will not divert a ship from its most expeditious course into deep water, if that were possible, simply to comply with conditions for mid-ocean exchange. What options are available in these circumstances? The internal stiffeners, particularly in DB wing and centre tanks, may trap sediment which is known to be a major habitat aboard for unwanted aquatic organisms, which may later be discharged with ballast water. Does the proposed system provide sufficient turbulence and scouring effects to ensure sediment is removed during the flow-through procedure? At IMO DE40, where the safety aspects of mid-ocean exchange were considered, there seemed to be uncertainty in some quarters as to how serious structural overstressing might be. An ex ship's master indicated that when he was at sea ballast water was taken on and discharged in mid-ocean without any such consideration. Therefore, a communication problem appears to exist between the designers of ships, the class societies, and the operators, although other delegates were clearly concerned that a problem with midocean exchange did exist, but the exact parameters were not known. The consequences of getting it wrong at sea might well be found littering the seabed. Dr G Armstrong (Three Quays Marine Services, London) Mr Lankester is quite correct in concluding that


flow-through exchange of ballast would not be appropriate on voyages which do not include a deep ocean passage, although having the tanks on continuos flow-through would minimise the transfer of non-native species. If flow-through could not be used, consideration would have to be given to the other methods described in the IMO Guidelines. Retention of sediment on internal stiffeners is a problem with tank flushing, and methods by which this can be minimised by design are described by Taylor (Ref 3 in the paper). The proposed system could not guarantee to remove all traces of sediment, but as there are approximately equal flow rates of water between all the floors, it is considered that there would be sufficient movement to minimise sediment deposition, without the risk of damage to the tank coatings due to scouring. I cannot accept that there is a communication problem amongst classification societies, designers and operators concerning longitudinal strength. Calculation methods are very well established and it would be naive to suggest that longitudinal strength is not an issue. Dr G Reynolds (Lloyd's Register) You referred to the research which LR undertook on behalf of the Marine Safety Agency into ballast water treatment. It may be of interest to note that following the work that you cited, laboratory research was undertaken into the effectiveness of UV radiation and silver/copper ions for the killing, or at least inactivation, of various organisms of concern, with respect to their carriage in ballast water. The findings of the research were that UV radiation appeared to reduce significantly the viability of the bacterial, phytoplankton and zooplankton species tested. The results for dinoflagellate cysts and diatom testing stages were, however, unclear. Although the former results were encouraging, further work would be required prior to any recommendation for the use of UV radiation for the treatment of ballast water onboard ship. The copper / silver electrode system tested in the project appeared unlikely to be effective for the control of biological organisms in ballast water. Dr G Armstrong (Three Quays Marine Services, London) I should like to thank Dr Reynolds for her comments, which give the latest findings of LR's investigation into methods of ballast water treatment. Capt J Isbester (Eagle Lyon Pope Associates, London) Ballast exchange by the flow- through method, as we know it at present, is a most unsatisfactory process. Even when three times the volume of water has been pumped through the tanks, as we have been reminded tonight, a 100 % change of water is not achieved. Overflow through airpipes is an unsafe practice, the classification societies tell us, and overflow through manholes in positions where that option exists requires the considerable effort of unbolting and rebolting manhole covers. The continuous flow of seawater across the decks for a number of days interferes with work on deck and plays havoc with the ship's maintenance.

Trans IMarE, Vol 109, Part 3, pp 257­269

A flow-through system which eliminates these problems would bring real benefits, and Dr Armstrong's paper is to be warmly welcomed for that reason. I believe that it may offer benefits additional to the ones that he has claimed for it. In some trades sludge or mud in ballast compartments is a major problem. Ships which regularly ballast in the Scheldt or the Maas, for example, can accumulate 5-10 cm of sludge or mud in their ballast tanks, thereby increasing the ship's constant and reducing her cargo lifting capacity. The weight of that sludge can run into hundreds of tonnes in larger ships. If the system could be used to keep the ballast compartments sludge free a real saving for the shipowner will result. Some bulk carriers, though not many, are provided with dump valves for the rapid discharge of ballast from the topside tanks direct to sea. Dump valves can cause operational problems when their use is not properly supervised, but they also confer real benefits for the rapid discharge of ballast, when required. Could the overboard discharges in the proposed system be used as dump valves for the topside tanks? The Canadian Coast Guard recommend that bulk carriers in ballast, when visiting cold weather ports, circulate the ballast where possible to prevent it from freezing. Most bulk carriers are not able to circulate their ballast water. Could the proposed system permit the recirculation of ballast water within the ballast tanks? Finally, some questions about the operation and maintenance of the system. There may be a weight saving in the reduction of sludge, but what will be the weight penalty incurred by the installation of the system? Have any estimates been made regarding the cost of maintenance and renewals? To what extent will heavy pitching and rolling result in ballast tanks spilling their contents, as already happens to some extent? If the system is designed with pipe capacities matched to ballast pump capacities, what will happen as the ship gets older and the ballast pumps become less efficient? Will it be necessary to retune or to re-callibrate the system? Dr G Armstrong (Three Quays Marine Services, London) I am grateful for Capt Isbester's comment that the proposed system eliminates a number of problems with the present methods of flow-through. The benefits of dump valves to permit the rapid discharge of topside tank ballast are acknowledged but, as Capt Isbester states, there can be operational problems. Whereas it would be possible to incorporate dump valves, I would not wish to compromise the `fail safe' features of the proposed system which minimise the risk of any unintentional loss of ballast. For example, if the ballast pump stops with the ship-side discharge valve open, the tank level will fall only as far as the discharge pipe entrance.The system could be extended to

permit recirculation of ballast, and the internal tank pipework would improve internal circulation. In the case of Ormond, the additional weight of pipes and valves is estimated to be 42t if GRP is used for tank internal pipework, and 85t in the case of steel. Maintenance costs are difficult to estimate sufficiently accurately, but ballast pump maintenance would, of course, increase. It is considered that the proposed system would have no influence either way on tank contents spillage. With reference to the question of ballast pump efficiency, the vertical position of the entry to the discharge pipe is arranged so that the tank remains full under the design flow rate of one ballast pump. If the pump efficiency falls then the steady state level in the tank under flow-through will be reduced, but cannot fall below the level of the entry to the discharge pipe. If the pump flow rate is reduced, then flushing will, of course, take longer, but it would be better practice to maintain the ballast pumps than to redesign the system. A W McConnell (P&O Nedlloyd, London) I would like to congratulate Dr Armstrong for a most interesting and thought provoking presentation. It is clear that this subject needs urgent debate as maritime authorities worldwide are expecting pre-emptive action from shipoperators. If a voluntary code is not followed and owners do not encourage their Masters to empty and freshen ballast tanks on ocean passages and demonstrate that this has been carried out, then it is inevitable that statutory enforcement will follow. Dr Armstrong's proposals for modifying ballast systems in order to exchange ballast systems on passage are rather complex and very costly. It would be difficult, therefore, to justify this level of expenditure. Dr G Armstrong (Three Quays Marine Services, London) I am afraid I cannot agree that the system is complex compared to other options in the guidelines. Costs quoted are for the retrofit at European prices, and the additional costs for installation during first build would be substantially lower. Mr K R Dagnall (Retired LR Surveyor) It is assumed that the centre tank has its own overboard discharge valve, as it is considered that as shown there could be a possibility of contaminating the ballast tanks through the open-ended pipes in the ballast and centre cargo tanks. Dr G Armstrong (Three Quays Marine Services, London) Mr Dagnall is correct that the centre tank would have its own overboard discharge. Figure 6 in the paper is not clear, and the intention would be to have a similar arrangement to that shown in Fig 7.



13 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate


You might also be interested in

Microsoft Word - BW-scoping-study-NW-Europe.DOC
Microsoft Word - Silent Invasion Briefing.doc