Australian Submarine Corporation Home
| Project Title |
| "Submarine Lead Acid Battery Performance Model." |
| Industry Contact |
| Peter Tromans |
| email: patroman@subcorp.com.au |
| Glenn Bate |
| email: ghbate@subcorp.com.au |
| Moderators |
| Basil Benjamin |
| School of Mathematics and Statistics |
| University of South Australia |
| Mawson Lakes, SA 5095 |
| Tel: 08 8302 3084 |
| Fax: 08 8302 5785 |
| email: basil.benjamin@unisa.edu.au |
| Mark McGuinness |
| Reader in the School of Mathematical and Computing Sciences |
| University of Victoria |
| Wellington, New Zealand |
| Tel: 04 463 5059 |
| Fax: 04 4635045 |
| email: mark.mcguinness@vuw.ac.nz |
Company Background
Australian Submarine Corporation (ASC) is the builder of Australia's fleet of new Collins Class submarines. Since completing the construction of new submarines the company has re-focused - to become a service organisation dedicated to ensuring the RAN continues to have access to six capable, fully operational Collins Class submarines. This includes optimisation of the arrangements for through-life support, including ongoing access to key technologies.
Australian Submarine Corporation Pty Limited (ASC) was formed in 1985 to tender and then to design, build and test a new fleet of submarines for the Royal Australian Navy, the Collins Class.
The Collins Class is the largest diesel-electric submarine built, approximately 80m long and of 3,000 tonnes displacement. There are six submarines in the Class.
ASC employs approximately 900 people, of whom about one quarter is blue collar. There is a multi-disciplined engineering department covering a wide range of technologies including an integrated logistic support system designed to maximise the use and performance of the submarines in one of the largest and most complex projects ever undertaken in Australia. ASC maintains quality accreditation to ISO 9001.
ASC has its headquarters and principal manufacturing and overhaul facility on the Port River at Outer Harbor, in Adelaide, South Australia. ASC also has a strong presence in Western Australia, at the home port of the Collins Class. Full service support is provided to the fleet for all maintenance activities that are undertaken in Western Australia.
Overview of Investigation
Diesel electric submarines utilise a large flooded lead acid battery as a secondary energy storage medium. The battery is the only source of power for the submarine while it is submerged. The batteries are regularly charged from diesel electric generator sets that provide power for the submarine as well as charging the batteries. Operating the diesel engines requires manoeuvring the submarine close to the surface while extending a snorkel on a mast above the surface to provide air for the engines. This operation is known as snorting. While snorting the submarine is considered to be indiscrete and is extremely vulnerable to detection.
The performance of the lead acid battery as an energy storage device is fundamental in determining several key submarine performance parameters. These are:-
1. Indiscretion ratio (IR – ratio of time spent snorting to total operating time on a mission) at specified Speed of Advance (SOA – average of submerged speed and snorting speed)
2. Submerged endurance
3. Overall range
To assist in determining the impact of submarine design changes on overall performance of the submarine ASC are developing a software model of the primary energy systems on the submarine. This model will include the following submodels:
1. Main Propulsion
2. Auxiliary Loads
3. Cooling System
4. Main Batteries
5. Diesel Electric Generators
The model will calculate energy flows and losses for a defined configuration and mission profile of a submarine and determine achieved IR or endurance or range. This information is important when evaluating the impacts of design changes on the submarine and for providing advice on alternative operational scenarios to the Navy.
Description
While most of the submodels can be easily defined as simple parametric or dynamic models the Main Battery model is more difficult. The amount of energy that can be delivered from a lead acid battery is dependent on both the past and future discharge rates as well as the previous charge cycles characteristics. On top of this the electrochemical effects of plate sulphation and ongoing plate corrosion can impact the energy storage capacity.
Predicting the amount of energy a battery can deliver from 100% capacity at a constant discharge rate is simple. Calculating this in a scenario where the battery is continually being subject to discharge cycle depths of around 15-20% at varying discharge powers and charged at high power for limited periods is far more complicated. The proposed model must take into account the non-linear performance of the battery under both discharge and charge, dependencies between various parameters and the electrochemical processes which affect the available energy at different discharge rates.
The interface to the Main Battery model in the SPM will have two components:
1. Configuration data provided at initialisation once and used to define a particular instance of a battery, i.e. capacity at different discharge rates, internal resistance, open circuit voltage, maximum charge current, no cells etc.
2. Run data such as current discharge/charge power and evaluation period. The model will need to return state of charge relative to current discharge/charge power and voltage at the end of the current period. Some error handling will be required to handle situations were discharge/charge limits are exceeded within a time period.
The model will need to remember previous calculations as these will impact on the current calculation of capacity.
Proposed Analysis
Several static parametric type models are available (Hoxie and Leibenow model) that could be adapted for use in the Submarine Performance model. However, use of these models will limit the ability to make global changes to the main battery configuration easily as they need to be characterised from trials data of an actual battery.
ASC would like to determine what analytic/modelling techniques would be best suited to representing this type of system and what key parameters may be required to define the model. In addition, it would be useful to define the design of such a model if time permits.