Locked Re: Batteries starting to get weak - MMM - What to do -
Scott SV Tengah
Mark/Arno,
Arno pointed out one huge problem with using voltage as an indicator of SOC and a trigger to start charging. While lithium has a much flatter voltage profile than lead with respect to load, it still has a voltage drop under high loads. As mentioned, my bank drops into the mid 24v range when I'm running the bow thruster with engine/genset off. My load varies dramatically. We may be drawing only 60w but then suddenly the washer and dishwasher hit the heating part of their cycle and that is aligned with the misses using the microwave and we decide to run the water maker. Oh and the fridge compressors turn on. At that moment, we could easily go from 30w draw to 6000w draw. Additionally, you really need to monitor and act upon cell level voltage. You could easily have 26v at the battery/bank level but an imbalance has one cell at 4.2v, but unseen at the battery or bank voltage level. Keep feeding current and the battery is irreparably damaged.
Mark - you stated "I think this high discharge rate has a significant impact on the Voltage vs. SOC results." Based on this and our variable load, unless you are able to consistently let your batteries get to their resting voltage, I still think voltage is not useful for determining SOC.
In my opinion, using voltage is a shortcut to try to make things easier. I am all for simplification as there is less to go wrong. But the manufacturers and the decades of academic research on lifepo4 all point to using a coulomb counter (like a BMV-712 or equivalent) to determine SOC and the target charging voltages are based on their research. Moreover, if I follow Victron's recommendations and a battery fails, they are on the warranty hook. They are incentivized to give you good information. It may be that the manufacturers and the decades of research and are wrong and that blogger and the MV guy is right, but I am not willing to risk my five figure lithium investment and my six figure boat on conclusions drawn from a sample size of two. I am certain the blogger wouldn't be providing warranty replacements if I tell him that I followed his advice and my batteries failed. :)
I am impressed that people are willing to go to the bleeding edge to save money. When I first started talking about installing a mostly-Victron system that was designed to work together, more than a few said that it was too risky to use "explosive" lithium. That was just two years ago. Now people are willing to mix and match batteries/chargers/BMS and trying non-standard charging routines. But I wouldn't be comfortable doing that on my boat on our intended, in-progress circumnavigation.
--
Scott
2007 A54 #69
SV Tengah
http://www.svtengah.com
Arno pointed out one huge problem with using voltage as an indicator of SOC and a trigger to start charging. While lithium has a much flatter voltage profile than lead with respect to load, it still has a voltage drop under high loads. As mentioned, my bank drops into the mid 24v range when I'm running the bow thruster with engine/genset off. My load varies dramatically. We may be drawing only 60w but then suddenly the washer and dishwasher hit the heating part of their cycle and that is aligned with the misses using the microwave and we decide to run the water maker. Oh and the fridge compressors turn on. At that moment, we could easily go from 30w draw to 6000w draw. Additionally, you really need to monitor and act upon cell level voltage. You could easily have 26v at the battery/bank level but an imbalance has one cell at 4.2v, but unseen at the battery or bank voltage level. Keep feeding current and the battery is irreparably damaged.
Mark - you stated "I think this high discharge rate has a significant impact on the Voltage vs. SOC results." Based on this and our variable load, unless you are able to consistently let your batteries get to their resting voltage, I still think voltage is not useful for determining SOC.
In my opinion, using voltage is a shortcut to try to make things easier. I am all for simplification as there is less to go wrong. But the manufacturers and the decades of academic research on lifepo4 all point to using a coulomb counter (like a BMV-712 or equivalent) to determine SOC and the target charging voltages are based on their research. Moreover, if I follow Victron's recommendations and a battery fails, they are on the warranty hook. They are incentivized to give you good information. It may be that the manufacturers and the decades of research and are wrong and that blogger and the MV guy is right, but I am not willing to risk my five figure lithium investment and my six figure boat on conclusions drawn from a sample size of two. I am certain the blogger wouldn't be providing warranty replacements if I tell him that I followed his advice and my batteries failed. :)
I am impressed that people are willing to go to the bleeding edge to save money. When I first started talking about installing a mostly-Victron system that was designed to work together, more than a few said that it was too risky to use "explosive" lithium. That was just two years ago. Now people are willing to mix and match batteries/chargers/BMS and trying non-standard charging routines. But I wouldn't be comfortable doing that on my boat on our intended, in-progress circumnavigation.
--
Scott
2007 A54 #69
SV Tengah
http://www.svtengah.com