Yet another LiFePO4 battery and Inverter thread… Or how I learned to love AC-DC

After several years of research this is going to be a long, fairly detailed, lead in to the whys and wherefores of my inverter installation. This not meant to be a tutorial but just some highlights and simple explanations. This is most definitely an experiment only time will tell how well it all works out.

Teaser:


I’m going to start with an overview of the risks, challenges and choices. I freely admit I picked solutions from many different very knowledgeable people Nigel Calder, Will Prowse, Peter Kennedy, Jeff Cote and others many on Club Sea Ray. By profession I am an “Integrator” I assemble solutions from commercially available components and that is what I did here.

In the marine market Lifepo4 lithium battery systems are still quite new. And there is no simple one size fits all “drop in” solution, regardless of what some battery venders say, nor even from many marine industry professionals. There are any number of designs that are neither “right” nor “wrong” just different. Each have pros, many have significant cons or risks.

My situation:
1994 300 Sundancer, never had a generator. I could add a generator at around $15k+/- not really cost effective for 28 year old boat. I could look at used or rebuilt generator at around $6k+/- but then life expectancy or unknown condition is a concern. I also need a complete system since the boat never had a generator so muffler, hoses, fittings, exhaust port, water inlet, etc. Then there are the additional concerns of Carbon Monoxide at anchor, risk running overnight.

Given this I chose to create an inverter system that is the equivalent of an AC generator since I feel I can do this for far less than the $6k.

First I will cover some of the issues with attempts to “drop in” a 12 volt LifePo4 battery.

1. LifePo4 batteries can absorb a significant amount of current during charging.

a. Pro this means you can charge quickly if you can provide enough power to the battery
b. Con NO standard marine alternator is designed for 100% output for more than MINUTES. So a 100A alternator will start to heat up after a few minutes at full output, as it does it will become less efficient dropping to 70% or less. Also if it runs hot enough long enough it will kill itself. ($)
c. Result, special alternators, regulators and charge controllers are required to keep the magic smoke inside direct charging LifePo4. ($ & Complex)

2. LifePo4 batteries are expensive and very unforgiving of “abuse”.
a. Require protection from extreme discharge currents, over charging, excessive float voltage, thermal damage, cell imbalance. These things are managed with a combination of the charging controller and a Battery Management System (BMS).
b. The BMS will protect the battery continuously but the final protection is an automatic battery disconnect. A sudden battery disconnect will cause a voltage spike that have been measured to 87 volts or more on a 12 volt system. This can easily damage electronics and is a known killer of standard alternators so again this requires special considerations.
c. LifePo4 batteries can provide high currents but there are limits. At 12 volts invertors and starters will draw very large currents a 4kw inverter can hit 350 amps. LifePo4 batteries have discharge rates that are usually referenced as “C”. So 1C is 1 x the Capacity of the cell, a 100A cell with a 1C continuous rate means you can safely pull 100 amps until it reaches low voltage cutoff. The same cell may have a maximum discharge rate of 3C or 300 amp in this case. This again is where a BMS will come in to enforce these limits. Exceed the continuous rate and it will be allowed up to the maximum duration specified for the battery at which point the BMS will disconnect the load to protect the battery.

3. Non-battery specific issues at 12 volt
a. As seen above a 4Kw inverter could pull up to 350 amps. This requires very large very short battery cables to prevent cable overheating and excessive voltage drop to an inverter. This would require 3/0 AWG cables for 4 feet (x2), that’s a lot of copper.
b. Similarly charging at 12 volt from a 200A alternator would require 2 AWG cables at 10 feet (x2) still some very heavy wiring.
c. All these cables require short circuit protection and manual disconnect switches, at these high currents such switches, fuse holders and T type fuses get expensive.

Ok so how will I avoid those complexities or reduce the number of them?
In my specific use case I feel the modifications necessary to the existing 12 volt systems is too broad to deal with. 12 volt is a carryover from automotive designs. Doing everything at 12 volt just because it’s common has too many challenges. Many charging components would need replacing or modification to cover all the possible issues adequately. It can be done, if designed with a new boat it may even be practical, but not as a conversion. You just have to throw away too many perfectly useful components and replace them with ones costing four times as much.

Existing Systems:
Twin 5.7 260HP Mercruiser I/O Alpha I Gen2 drives. Both engines have a basic 60A marine alternator. Both have a FLA marine starting battery not deep cycle; simple, reliable and cheap.

Starboard engine battery powers the engine and primary bilge pumps and the anchor windlass. The windlass is a very large short duration motor load similar to a starter. Since the windlass is never used without the engines running there is no great advantage having it on the house bank. A Blue Seas ACR currently charges the house FLA deep cycle battery. This engine also runs the power steering.

Port engine battery powers the engine and emergency bilge pump. A power steering pump that does nothing. Mercrusier configures both engines the same, this pump just recirculates fluid back to the oil cooler and then the reservoir. It simplified Mercrusier’s inventory. A few boats may have a priority valve to allow either engine to provide power steering, but it seems a rare option. This unused pump will become important later.

DC System
I will retain the 12v deep cycle battery as is for all house electrical and electronics this will eliminate any need to protect the 12v systems from a lithium battery disconnect. This also meets the new ABYC TE-13 and E-11 recommendations for critical system power. I will remove the ACR but retain the 1-Both-2 battery switch on the house for backup should I need to charge off the starboard engine. Under normal circumstances this house battery will charge thru a 30 amp DC/DC charger from the lithium bank. This gives me the extended run capacity of the lithium bank without any significant 12 volt system modifications.

AC System
I previously upgraded my 120v/30A to 240v/50A service with a heavy duty 60A pin and sleeve marine connector on the boat. This upgrade included a new shore panel for the 240v with a marine RCD/GFCI protection meeting current ABYC standards. I also upgraded the galley range to a 240v unit and the water heater was increased from 6 gallon 120v to 11 gallon 240v. Remember higher voltages mean less current required for same power use. This is why so many 30A sockets end up burning as they get corroded and loose. It’s a 1938 design that we have outgrown as we have added microwave ovens, stoves, air conditioners, heaters, etc.

PART 2
*********
New Systems:
Batteries
The LifePo4 bank will be a 16 cell series configuration for 48 volt nominal output. Using 48v has the advantages of much smaller cabling requirements. A smaller transformer in the inverter per kw. And less stress on the inverter internal components such as the FETs which are the primary solid state components that control the power to the transformer. This is all due to needing only ¼ of the current flow. A 4kw inverter at 48v only needs 85 amps vs the 350 amps at 12 volt.

This is a significant design choice and a fundamental problem with 12 volt high amp systems. Any resistance in any cable, connection, fuse, disconnect switch or component will be transformed into heat. Just 0.001 ohm at 350 amps will produce 122 watts of heat. At 85 amps will produce just 7 watts of heat. Any inverter over 2Kw you should be considering 24v, over 3Kw 48v really is a must.

The battery made up of LittoKala LifePo4 105AH prismatic cells configured in 16 cell series. This is just over 5Kw capacity. These are smaller than I would have gone with but this is an experiment and the batteries are the easiest component to damage. Also the price was excellent at $439 total with shipping. Although I ordered back early August 2021 and they finally arrived October 25th. If all works well and capacity is the only issue I will upgrade later as needed either more 105Ah for 16s2p (10kw) or just replace all 16 with 280AH (13.5kw) cells.


BMS
Heltec BMS\balancer with dual external disconnect relay/contactor (no FETs), thermal protection, Hall effect current sensor, eight temperature sensors, CAN, RS485, Bluetooth. It supports small pre-charge contactor, 200A charge and disconnect contactors.
This was selected and ordered in November 2021 and took 49 days to deliver arriving January 4, 2022.

Inverter
MPPSolar Hybrid LVX-6048 6Kw split 120/240, with 100A 450V MPPT charger, 120A Utility charger. LVX-6048 Inverter was my initial choice. Then a Growatt 6Kw 120/240 Split Phase unit has become available with 80A 250V MPPT, 60A Utility charger that has a better foot print and slightly cheaper.
Both the above have an integrated MPPT charger, but since part of my design is very “experimental” I ultimately decide on a simpler inverter/charger by Genetry Solar 6kw and a separate MPPT controller.
A high PV voltage is needed to take best advantage of the alternator output explained below most basic model MPPTs only support up to 95v. 6Kw is a bit over kill for my boat but I do want to easily start and run the Air Conditioning., typical load will be well under 3kw, but motor starting loads can be much higher.

Charging System
This is the unique part. Hydraulically driven alternator approximately 3Kw output.
Based on a Dynaset hydraulic generator motor and constant speed flow control, using modified 24V 70A alternator, external AVR and 3 phase rectifier. Expected output is 150VAC to DC at possibly 25A.
Whenever the main engines are running I will have somewhere between 2 to 4Kw of available power WITHOUT discharging the batteries. So I can run the AC to pre-cool the cabin underway and prior to anchoring. The existing power steering pump and oil cooler will be utilized. The pump requires one minor modification removing its internal flow control function, which is simply drilling out a flow restriction port.

I originally intended to replace the Port power steering pump with a Permanent Magnet Alternator. A PMA designed for wind charging. PMAs have no brushes so they would be intrinsically “Ignition Protected”. PMAs are rewound for higher voltage output at low rpm (wind) and they are most often built from modified 12v Delco 12SI units a relatively small frame. The units were limited to a maximum 2700 rpm. So with a pulley sized to prevent over revving the alternator at engine WOT it was going to be very low output under 1000 engine rpm. Other issues were a PMA having no brushes or field winding has no voltage regulation; the faster it spins the higher the voltage but at low rpm low voltage and low output. I wanted a solution that would charge well when just cruising around at displacement speed and not over heat/over charge at higher speeds. And the PMAs are priced quite high starting around $400.

After much consideration and after going with the hydraulic drive it became simpler and cheaper to mount the unit in the enclosure with everything else. This precluded the need for “Ignition Protection” as nothing inside the enclosure is IP rated. So I was able to change to a large frame 24v 70A standard brushed alternator for only $97 new. The larger frame and fan will run cooler at the outputs I desire. I removed the internal voltage regulator and diodes they can’t handle the desired high voltages and restrict air flow. I am rewinding the stator coils. ($50 magnet wire). I have purchased a universal external AVR (automatic voltage regulator). The AVR was $30,

The Enclosure
Mine is a gasoline I/O boat, a major issue was where to mount all the equipment. A generator would be mounting athwart ship just forward of the engines. The only space near large enough in the cabin of a 300 would have been to pull the water tank and put that in the engine room. The water tank is in a void under the aft cabin bunk. But I really don’t want all this heat and equipment in the cabin. Ventilation, noise and safety there would have been a big challenge.

My final solution is I located a stainless steel NEMA 4x (water tight, corrosion resistant) electrical box. The box is 36”wide 30” tall and 17” deep, picked up used for $250 with shipping. This fits perfectly were a generator would be. Being completely sealed it meets the USCG/ABYC requirements for separation of non-IP devices from fuel sources. It also should meet ABYC TE-11 proposed rules for the battery containment (steel box). It will be insulated and cement board lined inside this will contain any battery electrical fire and protect the components from engine room heat during normal operation. All wire openings will be water tight glands. It will be ventilated with forced inlet air blower (positive pressure in the box when running). The inlet and exhaust ducts will enter/exit at the top off the box be 6” flexible aluminum and be located separate from the engine ventilation.

I will do an extensive post on just the enclosure and the various safety items I will be incorporating. Some highlights: Passive fire protection, Active fire suppression, Remote emergency battery disconnect (electric), external emergency battery disconnect (manual), Remote BMS audible alarm,

Others will chime in I’m sure and have many valid and some not so valid opinions.
Let the games begin…

The enclosure is lined with 3/4" gypsum board for insulation and then 1/4" Harde cement board for fire resistance. The bottom has an extra layer of 3/4" closed cell foam padding to provide a bit of a cushion for the battery pack. The insulation is to protect the components from engine room heat and protect the engine room should an internal fire erupt. The black is 11ga steel plate bolted to the four corner mounting posts. The cutouts are to lighten it a tad where not needed for component mounting behind the batteries and under the inverter. The first fitting of the battery pack and fixture. These are 105Ah cells, I can add a second row or upgrade to 208AH in the same space and fixture.

The batteries, BMS and control contactors and Hall current sensor.

My god man….. where do you find the time?…. I am sitting here thinking I am screwed for an April launch….. and I am paying for rub rail up compound polish wax and nothing is broke except a door knob and I am thinking I can do that in the water

The hydraulic motor bell housing and alternator . The bell housing is for driving the "pump" (motor) with a gas engine. I cut it down and with some plate aluminum am fabricating a mounting assembly. Still have to TIG weld this when the weather breaks. My main shop is not heated. Doing most of this in my garage right now.

Cut the bell housing in half as it was to deep for the coupling.
fitting up for final depth and squareness.
Used one of the 1/4" aluminum plates as a true surface for sanding the housing to perfect fit.
Front plate drilled and bored for the alternator
Drive coupling also shortened and shimmed for fit to alternator shaft
Fitting up
Complete fit as will be welded

Blueone said:

My god man….. where do you find the time?…. I am sitting here thinking I am screwed for an April launch….. and I am paying for rub rail up compound polish wax and nothing is broke except a door knob and I am thinking I can do that in the water

Click to expand...

Time isn't a problem for this right now I can work inside, it's the budget.
I have a "to do" list for exterior too as as soon as the weather breaks a bit more, cant wait for Daylight Savings Time (March 13). And today was a tease here it was 29 last night and peaked at 65 already cooling down now. But when its warm this time of year it starts to rain which also sucks...

Great thread!! Love the move to 48VDC for the inverter! Surprised more folks aren't doing that. Not only does it minimize the AWG for the two cables for power, but it also minimizes the AWG for the SAFETY GROUND/BONDING cable that should also be installed for inverter systems. Lots of folks neglect that one (even installing one at all!).

km1125 said:

Great thread!! Love the move to 48VDC for the inverter! Surprised more folks aren't doing that. Not only does it minimize the AWG for the two cables for power, but it also minimizes the AWG for the SAFETY GROUND/BONDING cable that should also be installed for inverter systems. Lots of folks neglect that one (even installing one at all!).

Click to expand...

Good point on bonding, yes I will be bonding the entire enclosure since it is steel, as well as the battery ground side. I'll need to call that out on the schematics.

I love the idea of using the redundant P/S pump to run the alternator! Are you using any polymer "spider" between the coupler halves?

Nater Potater said:

I love the idea of using the redundant P/S pump to run the alternator! Are you using any polymer "spider" between the coupler halves?

Click to expand...

Yes, if you look closely at the photo with the pump and alternator fit without the bell housing you can see the tips of the black flex center. I'm being very careful to get the bell housing and plate perfectly parallel. But the coupling allows for a slight misalignment.

Novel project Patrick. Just made the time to read in its entirety. Looking forward to it operating. Here, if I can, are several comments and recommendations to avoid pitfalls.
From a standard automotive alternator perspective, it takes about 1 engine horsepower to drive 26 amps. So if you want a capability to drive 100 amps (maximum) you will need around 4 horsepower demand on the PS pump and pulley drives. Then there are the inefficiencies by driving the PS pump, the PS pump itself, flow restrictions in the tube and hoses to the hyd motor, the hyd motor, 24 volt energy, to finally 48 volt energy. Not hard to calculate but I'd guess for this test setup you are looking at between 65% to 70% (just an educated guess). It would be awesome to flow control the hydraulic fluid so the alternator RPM is always in its sweet spot which should be right around 4000 to 5000 RPM. I would look at the flow curve of the hydraulic motor to make sure it can achieve the RPM and required HP with margin; the system inefficiency needs to also be factored in here. Energy to heat will be a factor for the hydraulic fluid; you will need a cooler for sure; again not difficult to calculate and size but it will be needed.
All inverters especially large ones when first connected to a lithium battery bank have huge inrush current (literally thousands of amps) to charge the caps in very short timelines (a few milliseconds). This inrush will trip the BMS every time which you never really want to do. You will need a small resistor in the switching circuit so that initial inrush goes through the resistor then direct to the batteries. Some just short a resistor before switching; it is an impressive spark, arc welding level.
There are upsides and downsides to higher voltages in lithium configurations. For critical electrical service (like mine) either stay with 12 volts and a lot of call it "P's" and separate BMS or step up the voltage and greatly upsize the battery bank. So for others, Patrick is setting up a non-critical battery system; in that, should the battery trip offline or a BMS fails you don't loose the capability to safely operate the boat. He just looses his 240 volt AC power; much like a generator shutting down on a fault. So, in this configuration there is a single BMS to manage the 16s2p battery bank (that means 16 batteries in series and each of those series banks tied together in parallel). If, for example, he looses one cell, it fails, that entire 16 cell bank is kaput; he is at 50% capacity on a cell failure. I know this is a great test but in the final configuration you might want to look at a different battery configuration and multiple BMS to get the electrical system reliability up.
Lastly, the prismatic cells need to have some separators installed between the cells. That shrinkwrap is thin and any holiday will badly take down your bank; those outer metallic cases are the negative on the battery. There is nothing a BMS can do to save a battery if this happens. I use a thin FRP electrical isolator sheet from McMaster-Carr.
Lastly, look if you need this 48V system isolated from the boat's grounding and bonding system. There is a lot of discussion on that isolation in the Solar and Wind sailing blogs. However, if your inverter is integrated with the boat's AC power system it may all have to be bonded....
Tom

@ttmott an accurate analysis of the situation.

The BMS supports pre-charging with a third (smaller) contactor and resistor. I do more detail on each component as I get into the configuration phase.

Admittedly not particularly fuel efficient energy conversion from gasoline to charging electricity. But since it is just going to be a parasitic load on a 350 V8 not very important overall. Not like I will be running a V8 solely to charge the batteries. The expectation is charging will be during normal cruising. I will likely add an automatic field shutoff below 1000 engine RPM. So as not to affect maneuvering/idling.

The hydraulic flow control is set for 3600rpm (since the Dynaset is a two pole AC 60hz generator design for vehicles or heavy equipment). I have run a number of flow calculations for a GM Saginaw PS pump and this hydraulic motor. I will need 11GPM at about 1,500PSI for approximately 3Kw full load. Time and testing will tell what I can get at a given RPM. Power steering has a water cooler, will have to see if that is sufficient to keep the oil at an acceptable temperature. Will also be replacing the factory pump tank with one designed for remote reservoir, these are common for hot rodding. The remote reservoir will be 2 or 3 times the capacity off the little on pump tank.

I only expect the 60A out of the MPPT. But like solar any time I am not using the full 3Kw the difference can go into the batteries. And most things are intermittent like the Air Conditioning, the batteries will take the big hit to start up, then the alternator should be able to carry the load. While putting a little bit back in the batteries. When the AC cycles off, most of the alternator capacity will go to charging. Wash, rinse, repeat.

Only testing will really show how well it can work.

TE-13 and E-11 both require that the all batteries be bonded, no exceptions listed. But intended to anyway as mention above. Nothing 48v is planned to leave the enclosure, so minimum risk of cross shorting a 12v component to the 48v. There will be a few external communications lines for monitors panels.

The fixture and the components using plywood is for the initial mock-up and proof of concept. The plywood will eventually be replaced with aluminum once all the dimensions are finalized. For now there are cardboard separators in there and the end plates have a piece of the cement board between the cells and the wood.

There will be a Acrylic top cover after I fabricate the BMS harnesses. And then a plate to mount the components, photos of that shortly.

I wouldn't get too fancy on that resistor to charge the inverter caps. You must have a disconnect and fuse for each battery (for the others don't confuse Battery and Cell). The BMS disconnect can't serve as the manual battery disconnect. Blue Sea Systems makes a cool two step battery disconnect that can be configured that it first passes through the resistor then to the battery. Simple and error proof. Once that inverter cap is charged it will hold its charge at least for a couple of weeks depending upon the quality of the inverter build. This is that switch; I think it will work just fine at 48 volts at the amperage you desire -
https://www.bluesea.com/products/5511e/e-Series_Dual_Circuit_Plus_Battery_Switch

@ttmott
I updated the Viso for the Pre-charge and moved a few items for clarity. I may as well take advantage of the features that are available in the BMS, it's programed so it will always pre-charge before the main contactor.

But I will only have one battery, and I have a 2 pole DC 125A Circuit Breaker that is the manual disconnect.
There is no need for another manual disconnect just for the inverter, the disconnect just needs to be within 36" of the device. Since everything is within the enclosure it's all within reach.