In 2014 we published A More Flexible Power System for Dirona. This article ran in the April 2016 PassageMaker issue and aspects of this design have been offered as options on new Nordhavn builds. Since that article was published there has been considerable progress in lithium-ion battery technologies. All forms of Li-Ion battery technology offer considerably higher power densities than the old faithful lead acid batteries used in most Nordhavns, including Dirona. But, some of these Li-ion battery technologies used in laptops and mobile devices have suffered from fairly high profile failures leading to explosion and fire. However, the lithium iron phosphate (LiFePO) variant is a stable and far safer Li-ion battery chemistry, with large sales volumes driven by the electric car market in China.
Several leading marine manufacturers including Mastervolt are now offering these batteries for the marine environment, but it’s still very early days. Building a reliable and economic marine system requires detailed research and a careful design based upon multiple fail-safes. Mark McGillivray, owner of Nordhavn 5002 Panacea adopted some of the approaches covered in the More Flexible Power System for Dirona, but Mark also wanted to move Panacea to the more modern LiFePO battery technology. Others have made this move but what caught my interest in Mark’s work was the depth of his research and attention to both safety and detail in the design he has implemented on Panacea.
This system has now been in operation for nearly half a year and it’s performing well so I asked Mark if he would be willing to write up what he did to share with others. It’s a very nice design and I expect that we are seeing the beginning of what will be a slow industry change from Lead Acid to Li-Ion based battery solutions.
Although Carol and I were raised at the end of a fiord open to the wide pacific and had plenty of boating experience, when we purchased Panacea in September 2015, we were certainly ignorant in respect to trawlers. We cannot thank the Nordhavn Owners Group enough for the unselfish and generous assistance from the tremendously skilled and experienced Nordhavn owners and Expert Members. Special note however, must be made of the support that James Hamilton has provided us: not only is Dirona‘s website a treasure trove of information but somehow James finds time to respond to our unending questions.
Projects on Panacea that were more successful thanks to this support from the Nordhavn boating community include: installing an isolating transformer, upgrading the navigation systems and layout of the Pilothouse console, decreasing the quantity of the infamously fragrant black water hose by about 75%, replacing the VacuFlush toilets with Tecma toilets with gravity feed to the black water tank (no low spots to hold water), abolishing the requirement to clean shower filters, installing an NMEA 2000 network throughout the boat, implement the beginnings of a full Maretron system, remedied the engine room overheating problem by installing Delta T intake and exhaust fans, installing of an Ulrtasonic anti-foul system concentrated solely on the keel cooler area and a dripless shaft gland.
But the mother of all projects was one that was more important than all the others combined. When we purchased Panacea her lead acid house batteries were using over 3 litres of water per day. This was releasing over six cubic metres of a highly explosive mixture of hydrogen and oxygen gas into an enclosed lazarette. And people say lithium batteries are dangerous!
Instead of simply replacing the offending deep cycle batteries, we immediately discarded them and, following Dirona‘s lead, we embarked upon the journey to engineer “a more flexible power system” with a twist: using lithium instead of lead for the house batteries.
The decision to go to lithium was prompted by the following lithium advantages:
- Much more usable power for less than half the weight and space.
- Longer battery life, both in calendar year and cycles
- High charge and discharge capability: C rating. This equates to much less generator run times.
- Charge efficiency with less energy loss in waste heat.
- Flat voltage charge and load curves
- The ability to interrupt a charge without jeopardizing battery life: partially charging lead acid batteries leads to sulfation, a major cause of premature battery decline.
A more in depth article discussing these advantage is at https://www.victronenergy.com/blog/2015/03/30/batteries-lithium-ion-vs-agm .
Note that lithium phosphate, LFO, and lithium cobalt, LCO, batteries are very different1:
|Lithium Iron Phosphate – LiFePO4||Lithium Cobalt – LiCoO2|
|Energy Density||Lower 90–120Wh/kg||Highest 150–200Wh/kg|
|Max Charge and Discharge Current||Panacea‘s Lithium Batteries: 3*C rating||<= C rating|
|Risk of thermal runaway||Low: >270°C, strong oxygen covalent bonds inhibit thermal runaway 2
Over charging risks battery death
Full charge promotes thermal runaway
|Expected Cycle Life||>5,000||500 to 1,000|
|Nominal voltage||Panacea‘s are 3.28V (3.28v x 4 = 13.12V)||3.6V (3.6v x 4 = 13.24v)|
|Chemistry Environmental impact||Low – no rare metals||High|
|Rate of capacity storage loss||3%/month, Lower than LCO||Higher than LFO|
Empirical values are difficult to portray when various sources provide differing data depending on: chemical formulations, when an article was written, varying battery management parameters some from lab testing and others from real life, different manufacturing processes used, etc. The following is therefore simply a graphical representation of my interpretation of the major differences between Lithium Iron Phosphate (LFO) and Lithium Coboalt (LCO) based on the batteries we used on Panacea, our implementation and management, as well as my subjective interpretation of information I have gleaned from my readings.
Important considerations when dealing with LFO batteries include ensuring that:
- Overcharging does not occur.
- The voltage does not drop too low.
- Batteries are not held at a high charge. Unlike lead acid batteries, lithium batteries prefer storage at 50% SOC. Do not float, but let them continuously cycle between charge and discharge.
- Temperature is regulated. But certainly for us and the design we have implemented and in our present environment, this is a non-issue.
One aspect that took me some time to assimilate was the differing charge and discharge voltages in relation to SOC. While discharging (at 0.5C) our batteries are essentially full at 3.30 volts:
However while charging this same 3.30 voltage indicates that our batteries are quite discharged:
Understanding this, and the effect of low charging rates especially while under loads, further obscures assessment of the actual SOC necessary when configuring and fine tuning the charging and discharging of lithium batteries.
Our Lithium Project
Lithium batteries are not lead acid batteries and, if upgrading from lead acid, a thorough redesign of the battery and charging systems is required including separating the charge and load buses. By now it should be apparent that if we were to upgrade Panacea‘s house bank to lithium it would be appropriate to incorporate this into our project of “a more flexible power system”.
Having a desire to cut costs, design for cost-effective upgrades and ensuring that I did not put boat or persons at risk, we went down the custom path. We chose a renowned battery manufacturer and a respected BMS manufacturer that has English as a first language.
Panacea‘s lithium project included purchasing:
- 16 x Winston WB-LYP400AHA prismatic cells giving 1600 Ahrs @ 12 volts
- 2 x Lithiumate™ Pro BMS master BMS units
- 2 x 600 amp current sensors
- 4 x 500 N/O amp relays
- 2 x 400 amp fuse blocks
The basic design is:
A very important aspect of this design is the ability to lose an entire battery pack while still retaining 600 Ahrs available power. However, and this cannot be over stressed, this redundancy introduces a potentially severe consequence: once the 2 packs become disconnected, they must never be reconnected unless they are at an identical state of charge. This is because of lithium batteries capability to exchange such tremendous amounts of power: they will destroy themselves by trying to reach this equilibrium in too short a time period. Our design has 400 Amp fuses installed that will protect against this self destruction. I do not know of other BMS designs but ours is not capable of limiting this power distribution so this must be done manually.
Be aware that BMS stands for:
- Battery Monitoring System and
- Battery Management System
It is important to note the difference as you will require a Battery Management System that automatically manages the lithium battery packs. The monitoring page of our Elithion Lithiumate software is depicted here:
The other tabs of Lithiumate provide the tools to change parameters depending on the type of lithium battery as well as other environment specific issues or limit the usable capacity parameters to improve safety longevity of the packs.
In terms of expected lifespan, if we still own Panacea in 10 years, I will be disappointed if any of these batteries require replacement. We do not push the batteries up into their charge and discharge knees, and because of this I expect a much longer life than even the manufactures are touting.
We have 2 Lithium battery packs, one on each side of our lazarette. The following picture is the top of one pack with the cell communication cables and connecting cables attached:
The following shows our BMS, 2 x 500 Amp relays, fuse block and 600 amp current meter.
The cable in the top left emerging from the hole in the battery box connects the lithium batteries to the 400 amp fuse. The other side of this fuse connects to the bar on the right that holds the 600 amp current meter. The charge and load relays are connected to the other side of this bar. The feeds from these relays go to the charge and load buses. The front of these connections are protected by 6 mm perspex: you will see a round cutout to gain access to the 9 pin serial connector. The 2 BMS’s are connected to a PC in the Pilothouse via a serial cable that has a USB converter before the PC.
Because lithium batteries will take all that an alternator has to give, it does not get a rest as when lead acid batteries go into absorption phase. The result is that alternators can easily overheat and burn out when trying to deliver lithium batteries demands. Also note that, lithium batteries do not like the absorption phase as this holds the voltage high which shortens their life. Therefore, we replaced our regulators with Balmar MC-16 and set the maximum output to 80%.
For anyone keen to implement their own custom lithium battery system or simply to learn more about this technology, I highly recommend investing the time to read through the following articles:
I have spent many hours reading and studying these pages and, even today, I still go back and check our implementation against this content. I do not suggest that you must follow their opinions blindly. I disagree with some of the views expressed, but they are valuable resources nonetheless.
We use Elithion’s Lithiumate™ Pro BMS Masters and are very happy with them and their support. However, we did not do enough homework beforehand and purchased 3 Chargemaster 100/12Vs that these BMS cannot talk to. We have it working, and I am happy with it. However, the path has been more arduous, complex and expensive than if we had we picked a dumb charger tested to talk on the Elithion CAN BUS.
Undertaking a lead acid to custom lithium conversion is not for the faint hearted. Pre-packaged solutions by companies like Victron and Mastervolt may be the safest, and possibly the cheapest way to go for those who chose to spend the time sailing over research and development.
Would I do this all again, knowing what I know now? In a heartbeat, although I would change a few things:
- Involve any skilled labour required in the planning stages, defining every piece and documenting every step before purchasing anything.
- Increase the number of batteries by 50%. They would still weigh less that the batteries they replaced yet give us even longer run times.
- Purchase the retaining jigs and straps from the battery supplier instead of manufacturing our own.
- Use shielded serial cable for the BMS control and monitoring cable that runs to the pilothouse.
- Instead of using Mastervolt chargers I would investigate the use of basic chargers that talk to the CAN bus of the BMS so that the BMS was in total control. However, because we have implemented a redundant BMS solution, this may not be a simple amendment: how do 2 independent BMS’s synchronize to control the chargers on the common load bus? If the choice is between our redundant BMS setup or a non-redundant system that is simpler to manage, I would probably stay where I am.
2016/1/9 Update from Mark McGillivray:
If the question is would I do the upgrade to LiFePo4 again, yes in a heartbeat. Saving weight and space as well as having to run our generator only once a day on anchor is heavanly.
However there are a few issues that I still contend with:
- Maintaining a wet cell (lead acid battery) at full charge is fairly simple and straightforward and leaving it at full charge is prefered over leaving it partially charged. Lithium formulations do not like being left at full charge, prefering to sit at 50%. If the operational paramenters of a lithium based system were this simple, it would seem reasonable to float the battery at 50%, say while connected to dock power, until you wish it’s full storage capacity as when you are at anchor and need to charge via generator. Even then, maintaining lithium batteries at 50% SOC is not an easy task as lithiums charge and discharge curves are very different and very flat. Our BMS calculated SOC but it needs to be recalibrated constantly and it’s SOC accurancy is always in question
- An article at http://batteryuniversity.com/learn/article/bu_808b_what_causes_li_ion_to_die maintains that charging at too low a voltage causes SEI buildup on the anode. This article is mostly centered on Lithium Cobalt formulations, therefore I have not had it confirmed that this SEI buildup also happens with Lithium Phosphate formulations. If this also applies to LiFePo4 batteries then this could be a good reason not to float.
- When on an extended voyage, the charging system may keep the batteries fully charged. How detrimental this is on a say 3 week voyage, I do not know. This is probably not a significant issue, but it is something I ponder. While on an extended voyage, one option to cater to this would be to “float” the lithiums at a lower voltage so that they are not held at 100% charge. Certainly I would not have the engine alternators keeping the lithiums at 100% SOC as that maintains them at a high voltage promoting cathode SEI buildup.
- For Panacea, on dock power here in Australia, we often prefer to run our 60Hz systems via our inverters so that our systems receive totally clean power at their designed 60Hz. We have 3 x Mastervolt Chargemaster 100/12Vs chargers that are capable of drawing a little over 15 amps. When plugged into a 15 amp dock receptacle, we have an issue when simply running only one of our 4 HVAC units as we can only pull about 12 amps before the shore circuit breaker trips. We therefore set our chargers to draw about 12 amps. As a result we can only run one AC unit for about 8 hours before we must turn it off to enable the chargers to replenish our lithium banks. Our work around for this problem is:
- To run our 230 volt AC systems such as air cons, dryer, washer, water heater and water maker on our 12 KW isolation transformer. This means that, in Australia, those systems run at 50 Hz not 60. If we only have access to a 15 amp circuit we can run 2 HVAC units.
- To run the remainder of our systems on inverted power such as 120 VAC microwave, fridges and freezers and all 12 VDC systems.
- If we have access to a 30+ amp power receptacle we are well set to run almost all systems off the isolation transformer or invert with the capability to run one HVAC indefinately.
My recommendation for anyone interested in a Lithium upgrade would be to certainly go with it, but buy a packaged solution from a recognised provider of systems such as Mastervolt or Victron and let their packaaged systems deal with all the issues under warranty. I would probably lean towards Victron as they seem to understand lithium better and I have been advised that their solution does not require separation of the charge and load bus, simplifying and reducing the installation costs.
About the Authors
Carol and Mark McGillivray hail from Port Alberni, Vancouver Island. After a couple of years training in Vancouver and Victoria they spent 5 years in the Canadian arctic and subarctic before immigrating to New Zealand in 1982. Carol’s qualification as a registered nurse entitled them to immigrate to New Zealand. Mark’s certification as a commercial diver gave him the opportunity to play with large eels in liquid mud. In 1984 they started importing, building and selling PCs. Business grew and evolved and today delivers clients complete IT services including full Managed Services through a custom built data centre situated in Palmerston North.
Mark and Carol purchased Panacea, a 50 Nordhavn, in September 2015 and, to quote Carol, “Mark has devoted most of the last 20 months on his retirement project boat”.
And, to quote Mark: “It has been a hell of a ride!”