Firefly Batteries [rec: no more than 4 strings be wired in parallel]

David Vogel

Hi Paul,

Thanks for your question. Like most things, it seems, there is anteresting back-story.

Also, a correction to my MultiPlus write-up: I actually purchased the Firefly batteries in late 2019, not the year before, as I mistakenly stated. (In error, I tied the events to the '19' in CV-19, but of course this happened in 2020 ... . So, just to correct the record, I ordered the FFs in late 2019, and they arrived and were installed in French Polynesia is February 2020, just as COVID was becoming a thing.) OK, back to the chase ....

To your questions:

Q: Did Firefly not advise you on the limitation of 8?
A: No, when I ordered the FFs in 2019, there was no limit to the number of series-pairs (more detail on this below). I went for 5x series-pairs, to provide 290Ah 'useable' capacity assuming a routine DoD of 50%, more than doubling the useable capacity I had previously with the 12x AGMs (assuming a routine DoD of 20%). I elected not to go for 6 pairs simply on the basis of cost (bang-for-buck). I chose not to go for only 4 pairs, as I was unsure of the technology, and wanted to exceed the doubling of the prior installed capacity, in order to make provision for some eventual degradation in capacity (which I have not experienced thus far, touch wood).

Q: Have you experienced any issues?
A: short-answer, no, not in normal use while we have been 'out there'. Long answer, it's a little more complicated than a simple yes or no, so please read on.

As part of background discussions with Bill Kinney a few weeks ago, re his experiences and configuration settings etc, I discovered that the Firefly manufacture OASIS had updated the User's Manual (at least) twice since the 2018 version I had been using as my reference.

Some internet sleuthing revealed a 2020 update, at which time the charge recommendation changed from "continue charging until the charging current drops to 1.5A for the G31", to become "Continue charging until the charging current drops to 0.5A..."; along with introducing the recommendation regarding the Periodic Fast & Complete Charge Cycle, which "should end with a 24-hour float charge at 13.5V for the G31".

Then, a further update in 2021, when we first see the recommendation: "no more than 4 strings of G31 ... to be wired in parallel"; along with other refinements, such as the SOC vs OCV curve.

The 2021 edition is attached. I will check the file-store on the forum and, if necessary, place all the editions there, for reference.

Further info, that may be of interest to other FF owners, actual or intending:

When I arrived in NZ November 2021 (Fireflies having been in hard use, remote-area cruising since Feb'20), I did a state-of-health check on the batteries, and they all tested fine. (Note: these tests are done using an on-board battery tester, which infers the state-of-health, rather than doing an actual full discharge cycle at 0.1C (or whatever), so the results are indicative only, not 100% conclusive.) Then, in early-November, as part of the preparation for the MP upgrade, I did another check, and was somewhat surprised to see that the indicated health had dropped off to ~70%. I was surprised because the batteries had, for the most part, been floating at 26.6v on shore-power for the previous 10 months, apart from an approximately monthly draw-down of ~120Ah (run-down to SOC80%), and re-charge. I had assumed that, as these batteries are essential lead-acid chemistry, holding them on float would keep them in the best possible condition whilst with access to shore power. While I have nothing definitive, subsequent anecdotal discussions with a few electricians (who have seen a number of sets of Fireflies of various ages and with various backgrounds) seem to indicate that Fireflies do not like being held on float for extended periods, but prefer to be exercised. The thinking being that the foam electrolyte somehow gets stratified and "lazy". This would seem to be in line with the manufacturer's recommendations for periodic deep-discharge followed by a high-current charge and above-float hold, basically to somehow blast the electrolyte out of its stupor. Anyway, I ran the batteries down to 50%, a little further, and the observed discharge curve matches almost exactly the SOC vs OCV curve published in the 2021 update to the User's Manual so, at the moment, I am not too worried that the batteries are irretrievably damaged.

That said, with the Multiplus now on-line (and the ability to quickly and easily tailor the charge profiles), I plan to do a deep discharge on each series-pair, and a fast-charge at 0.4C, and then re-do the tests. But there is a lot else also going on at the moment, all clamouring for attention so, in the interim whilst still on-the-dock with shore-power, I am no longer holding the batteries on float, but routinely discharging the whole bank down to SoC50%, recharge to SoC 80% @ 0.2C, and repeat. Then with a weekly full charge (to Charge Acceptance Rate <2.5A, previously being CAR<7.5A, representing quite a change). Then in the New Year, when things settle down again, will do the full discharge-recharge routine on a pairs-wise basis.

In the meantime, I would be pleased to hear any experiences or suggestions from other FF owners, or reference to other information that may be relevant.

Thanks, in anticipation,

Blue Skies,

SM#396, Perigee


On 10/12/2022, 1:49 am, "Paul Dowd and Sharon Brown" < <> on behalf of paul.dowd@... <mailto:paul.dowd@...>> wrote:

Hi David,

A very interesting and informative account, many thanks for sharing. I was about to have the Multiplus installed in the ER, but this seemed to present problems in locating it satisfactorily, so your post is timely. I was considering going for the Quattro, which is a lot more expensive, but will now consider the companionway locker. One question regarding your firefly batteries: I recently had these installed and in doing so was told by Firefly that I could not use more than 8. I see you have 10, so am a bit puzzled. Did Firefly not advise you on the limitation of 8? Have you experienced any issues?


Paul Dowd & Sharon Brown
S/Y Ya Fohi, Amel 54 #98
tel: +44 (0)7710 466619
skype: pauldowd
web: <>

-----Original Message-----
From: <> < <>> On Behalf Of David Vogel
Sent: 08 December 2022 21:23
To: <>
Subject: Re: [AmelYachtOwners] Inverter Upgrade - Victron MultiPlus

Hi All,

My inverter upgrade, augmenting the pre-existing 1,500W inverter (as installed as OEM) with a Victron MultiPlus (24/3000/70-50), is finally complete.

Below, documenting the generic system configuration and underling rationale/s, for the benefit of those facing similar upgrades in future.

The planning started shortly before acquiring Perigee in 2016, with the initial PV (4x 100W, rail-mounted) installed in MQ 2017; then WindGen in SXM, 2018 (Rutland 1200, top of mizzen mast, made easier when the masts were removed for replacing the standing rigging with ACMO to OEM spec); M2-1828 AC multi-meter ordered / delivered Panama, 2018 (faulty on delivery, replaced 2022 under warrantee). Victron Multiplus ordered when the legacy AGMs started to fail in French Polynesia, late in 2018. The Multiplus arrived NukuHiva, along with 10x Firefly batteries, in February 2019 . The batteries were installed immediately, but the Multiplus install has needed to wait for a more integrated systems approach (and an AMEL-friendly electrician), and to await a location with readier access to a suitable level of shore-based resources and support; detailed discussions with 4 installers took place along the way, before finally settling on a fifth, who assisted with the job, along with cleaning up the rats-nest of the DC wiring adjacent to the battery box (e.g. new +ve and -ve bus-bars, adjacent to the battery switches, and other aspects of prior solar and wind-gen installs).

Points-of-interest are:

Battery Monitoring: replaced the Xantrex Lite with a Victron BMV-712, which is networked via Bluetooth and VE-Direct; this works in combination with a Bluetooth networked Victron SmartSense which provides networked temperature sensing (and battery voltage sensing, as used by PV MPPTs).

Victron MultiPlus 24/3000/70-50. Chose VICTRON over Mastervolt - as did not see the need for cZone integration, as (perhaps, and arguably) this is the primary benefit of MV over VE; have a tendency towards the potential of open-source solutions as well. Then, chose the Multiplus over Quattro, as have taken an in-principle decision that "if it ain't broke, don't fix it", that is, to re-use as much as possible (in this case, the existing K1 / K2 transfer switches, and associated wiring & cable runs), whilst also minimising disruption to other existing/working equipment and systems. Adding 3,000W of inverting represented a doubling of the pre-existing 1,500W inverter (which has been retained, using the pre-existing K2 transfer switch, so as much as possible remaining independent of the Multiplus install). Also, chose a 3,000W (not 5,000W) unit, as a greater-capacity inverter would need higher-capacity batteries which, in turn, would need to be kept fed with higher capacity charge-source/s (solar). Similar rationale for 70A charger, not higher. Also, adopting an evolutionary approach, such that incremental improvements (such as increase in inverting capacity) does not drive or necessitate an upgrade elsewhere (such as Lithium). Holding to 3,000W will also (I hope) limit the consequence of the "use whatever you have" effect. Chose the unit with 50A transfer rating, over the one with only 16A capacity, so as to be able to passthrough the full output of the 7.5kVA ONAN GenSet (for battery charging, amongst other things).

Multiplus, installed in hanging locker adjacent to companionway. Rationale:
1. minimise length of DC cable runs;
2. avoid the heat of the engine room - having experienced heat-soak in the engine room, did not want add more heat sources there; 3. avoid having the Multiplus co-located with the other inverter and chargers (i.e. physically separate these pieces of equipment, if possible, to guard against broad-scale effects of an engine-room incident such as water-maker HP leak); and 4. reduce risk of heat-related reduction of inverter/charger efficiency due to high ambient temperatures.
Downside: extra noise in saloon when with high charge/inverting loads.
Offset: co-locate with new solar MPPTs, clear access to associated RCDs, CBs, distribution/fuse boxes, etc.

System configuration: via VictronCONFIGURE (for initial setup), thereafter VictronCONNECT (app on iPhone, iPad; and MacOSX via USB-C into ethernet/VE-Bus) and VRM (browser via internet, plus local).

Monitoring - Multiplus (and system): via Victron CERBO GX. Elected for the CERBO over an OCTO GX, for the more powerful CPU and better WiFi performance (despite the 10 VE.Direct ports offered by the OCTO; CERBO has only 3); use VRM app on iPhone/iPad and online, and B&G MFD at NAV STN (details below).

Control - Multiplus: Digital Multi-Control panel (DMC-GX), located below the ONAN control panel in the galley. Rationale: try and keep all the primary charging and inverting controls located within a single eye-glance and arms-reach. (The legacy 1,500W inverter has been held in situ, and still able to feed the AC system via K2 when the Multiplus is off.)

Monitoring & Control - Cerbo & System: via VRM (browser, and app) and HTML5, the latter achieved via ethernet to a B&G Vulcan 7" (V7) display. Elected to go this path in order to avoid the need for a separate dedicated display for the Cerbo (power, space & visual considerations); achieved by using the pre-existing B&G Vulcan 7" MFD, which has been panel-mounted (with only minor work required) to replace the retired B&G Hydra2000 display. The V7 also provides back-up/redundancy to the Furuno TZT14 chartplotter and integrated AIS functions at the NAV STN (but the V7 does not provide redundancy for the radar), while also providing enhanced wind information (waterfall history tapes) as per helm instruments (3x Triton2) - this is very useful when in inclement conditions, as can assess all parameters from below, at the NAV Station (note: all autopilot controls have been deliberately retained at the helm only). The CERBO is also connected to the N2K network (note: the CERBO's 12vDC N2K-supply fuse is removed, in order to avoid DC loops), meaning that electrical-system status and values (such as AC/DC voltages, currents, temps, warning & alarms, etc) can also be displayed on dedicated windows on the V7 MFD (and anywhere else where the N2K data-stream is available).

Other changes:
+ installed Victron battery balancers.

+ N2K network is newly separated into two independently-powered halves:
+ <MHU-to-DST-to-NAV STN+xxx> and
+ <Chartplotter-to-Autopilot-to-helm-to-GPS330B>. (All N2K data that is
+ available on one half, is passthrough to the other half.)

+ Furuno TZT14 Chart-plotter (and radar, plus the primary FA-50 AIS) 24V supply is powered by the main NAV STN circuit breaker via dedicated sub-bus; the 12VDC supply to the chartplotter, which powers the 'aft end' of the N2K network, is powered by the NAV STN SAILOR 8A 24-12VDC supply.

+ re-allocated previously unused capacity on the Sailor 8AMP "Always ON" 24-12VDC converter (previously FM-AM radio only, channel pre-set), now provides "Always ON" (but switched) 12VDC to the 'masthead half' of the N2K backbone; this power supply now also powers the CERBO, the B&G V7 MFD (which is also separately switched), and a (switched) 4G/WiFi Router. Doing so allows the CERBO to be monitored and controlled remotely (unless the Sailor 'Always ON' 24-12v DC-DC converter is switched off at the unit).

+ reconfiguration of the AC panel so that the pre-existing 100A and 30A battery chargers are powered only from the Multiplus AC-OUT-2 (AC-OUT-2 is switched only when AC-IN is available to the Multiplus, from either shore-power or genset, genset priority, as determined by AMEL OEM transfer switch K1). Thereby preventing an inadvertent charge-from-inverter downwards-spiral scenario, whilst also providing boosted battery charging capacity of up to 100+30+70 Amps (or combination thereof, max. 4,800W from 7.5kVA genset; lesser when on shore power, with power limiting applied via DMC panel).

+ Replacement of the analogue AC voltmeter (above the AC panel) with a Bluesea M2-1838 AC Multimeter; enabling greater visibility of AC loads at the point-of-control.

+ Installed VESPER XB-8000 (in SW controlled RX-only mode): provides
+ on-board MOB and Anchor Alarm, plus AIS to helm instruments (3x
+ Triton2) and V7 via N2K. Re-broadcasts the AIS and N2K data-streams
+ via WiFi for use by independent devices (e.g. iPads/iPhones running
+ iSailor, PC/MAC running O-CPN). Provides redundancy for the primary
+ Tx/Rx AIS (which is an ethernetworked Furuno FA-50 displayed on the
+ primary TZT14 chartplotter). (I elected not to go for the integrated
+ CORTEX solution, as I wish to retain AIS independent of VHF COMs
+ capability.)


Replacement of 4x 100W rail-mounted solar panels with 4x 310W panels (2 on the rail, one each port and starboard; 2 on bimini). Replacement of 1 x 100/20 MPPT with 2x 100/30 MPPTs; retaining the 100/20 thereby providing the optional election of retaining some of the existing 400W, or supplementing the new 1,200W with fore-deck or boom-tent mounted flexible / semi-flexible panels.

Future considerations:
Upgrade to Lithium: contingent upon aging capacity of existing 580Ah FireFly batteries, assumed timeframe remaining 3-5 years, at which time the presuemed costs of lithium with reduce in real-terms, as well as leveraging the benefits of flow-on technologies resulting from EV uptake, hybrid grid-storage solutions, systemic design considerations in marine applications, modularisation, interoperability, and institutional aspects (e.g., insurance), etc etc.

As usual, comments, questions, and suggestions all welcome.

Blue skies,

SM#396, Perigee
Bay of Islands, NZ

From: < <>> on behalf of Bill Kinney <cruisingconsulting@... <mailto:cruisingconsulting@...>> Reply to: < <>>
Date: Tuesday, 12 April 2022 at 12:03 pm
To: < <>>
Subject: Re: [AmelYachtOwners] Standalone Inverter as an AC IN source to a Victron MultiPlus


Thanks for the diagram... What you propose looks good!

Bill Kinney
SM160, Harmonie
Ragged Island, Bahamas

Ya Fohi - Amel 54 #98

Arno Luijten

We had 8 Fireflies on our 54 for about a year and found out they handle extensive floating quite badly. I’m not saying they are bad batteries but in our situation they were not performing as expected. We found the internal resistance of the Fireflies to be quite high at some point causing the system voltage to drop below 24 volts quite fast when a moderate load was applied. Maybe this was because we live on shore power most of the time but in the end we replaced them with LiFePO and that made things much better. We are now using 24Volt LiFePO batteries which is favorable over 2*12 Volt sets because of the BMS regulation.


Arno Luijten
SV Luna