Lynx Shunt 1000a Fire Hazard

Question

question

cookinwitdiesel asked Jun 21, '23

Lynx Shunt 1000a Fire Hazard

My dad's system at his ranch has now had dangerous high heat and smoke when running only about 700a across the ANL 750a Blue Sea Systems fuse in the shunt. The fuse did not blow but essentially otherwise cooked due to heat. All connections were snug and tight.

The first time, the plastic holding the busbars actually melted. We replaced the shunt and this 2nd time the fuse cooked (and again, did not blow, per VRM max current was right at 700a).

What is the best/recommended fuse for this setup that will not create a dangerous thermal bottleneck?

We are currently using this:
https://www.bluesea.com/products/5163/ANL_Fuse_-_750_Amp

Pictures of the first failure after which we replaced the shunt and then the 2nd incident with the new shunt. In both incidents, the fuse did not blow. VRM screenshot is the 2nd incident and the last 2 pictures.

fuses lynx shunt fire

315243908-840182160362049-7730592499707759671-n-1.jpg (283.0 KiB)

355420632-282734660822118-4981676582536193283-n.jpg (418.2 KiB)

315293964-1712137062490499-2850625929656513413-n-1.jpg (243.4 KiB)

355206143-650394330476182-7396755183211040959-n.jpg (99.7 KiB)

1687365462864.png (38.8 KiB)

6

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

janieronen commented · Jun 21, 2023 at 05:04 PM

That's serious overheating. Looks like that fuse is not only what have been overheated. How is load side busbar also heat damaged? What were environment temperature and is there sufficient amount of cooling airflow?

With 24V 700A is very strict about contacts. I have changed system voltage from 24V to 48V after noticing that cable and connection losses went absurd high with over 10kW power. As first unit were already overheated I suggest that in future fit few temp sensors to the bus bars to give alarm when rising close to cable heat limits. First test cycles heat camera is best tool but with probes you can be sure that there is no fire hazard.

Can you plot DC-power and AC-power to same to get picture how much energy were lost? I would be also interested did your AC-load increased after 19:00 or did cabling/busbars overheated and load increased because of higher temperature. Co-efficiency factor drops when overheating dramatically what causes even more resistance.

0 ·

screenshot-2023-06-21-at-195430.png (187.1 KiB)

cookinwitdiesel janieronen commented · Jun 21, 2023 at 05:18 PM

He does have temperature monitoring and probes now, added after the first incident. Seen temperatures as high as 250f multiple times. The bus bars got so hot they melted the plastic posts they are supported by the first time. This prompted a full replacement as we could not trust any of the physical connections on that unit after that.

0 ·

1687367885727.png (2.3 MiB)

1687367920744.png (2.0 MiB)

janieronen cookinwitdiesel commented · Jun 21, 2023 at 05:41 PM

Ouch. Cables normally have something like 150F/70C maximum temperature so system should automatically been shutdown at that point to prevent short circuit. My experience I would not use Lynx for over 400A continuous currents as limited heat dissipation often creates issues if not force cooled environment.

0 ·

cookinwitdiesel janieronen commented · Jun 21, 2023 at 06:04 PM

There was a good amount of loss, but that is not all heat in the conductors. There are 6 24/3000 MPs and 14x 250/100 MPPTs all powered on as well. Clearly a lot of heat was still generated though! My dad went down and bypassed the inverters and powered down the system when the smoke alarm went off, that is why the data stops. Smoke was from the fuse cooking.

We would love to have done a 48v system but the UL listed inverters were only 12 and 24v at the time.

0 ·

1687370643809.png (79.2 KiB)

janieronen cookinwitdiesel commented · Jun 21, 2023 at 06:27 PM

That is way too much current for a lynx if 1400A total from MPPTs is possible to push power to Multis and battery.

If remember correctly in the Lynx manual it's said that all connected load/generation fuses cannot exceed 1000A in total.

You should definitely to go with custom massive bus bars and physically larger fuses which can dissipate more heat. That's my own suggestion even love VE-products.

0 ·

cookinwitdiesel janieronen commented · Jun 21, 2023 at 06:30 PM

I do not push anywhere near that much current - honestly could have done 250/70s in this design. Only 23kw of solar panels (56x 415w) and max charging is set to 700a with DVCC as well as max inverter power set to 14400w with ESS.

0 ·

Modifications Space

Node Red Space

Answer 1 · 2023-06-21T16:59:41Z

natebert answered · Jun 21, 2023 at 04:59 PM

It looks like the fuse is the source of heat. Maybe you've got a couple from the same bad batch?

Or maybe, could the use of Bussman/Eaton's 800a ANN model offer additional surface area as the tabs on the ANN model of fuse appear to be thicker?

ANN:
https://www.eaton.com/content/dam/eaton/products/electrical-circuit-protection/fuses/bussmann-series-supplemental-fuses/supplemental-fuses-limiters/bus-ele-ds-2023-ann.pdf

vs
ANL:
https://www.eaton.com/content/dam/eaton/products/electrical-circuit-protection/fuses/bussmann-series-supplemental-fuses/supplemental-fuses-limiters/bus-ele-ds-2024-anl.pdf

11

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

cookinwitdiesel commented · Jun 21, 2023 at 05:04 PM

I agree, the negative bus bar which handles the same current shows zero signs of high heat. Only around the fuse on the positive side.

0 ·

Alexandra ♦ cookinwitdiesel commented · Jun 21, 2023 at 07:04 PM

So it is either not torqued correctly - which is a common problem install thing or the fuse is too closely sized to the system, other than the suggested bad fuse batch.

Smoke detectors for the win. Should be on must all installs.

Just a short comment on the "only 700A passing over".... that is not toy power. The install has to be on point in all the small details.

3 ·

cookinwitdiesel Alexandra ♦ commented · Jun 21, 2023 at 07:42 PM

Agreed, this is whole house power which is why we went with a high quality vendor for the parts (Victron) and designed it all out carefully. The "only" was in relation to the busbar ratings of 1000a indicating they should not be anywhere near their design capacity and therefore not a source of heat. All nuts were torqued to 14 n-m as recommended in the documentation.

In the US at least, pretty sure smoke detectors are mandatory for installs like this.

0 ·

janieronen cookinwitdiesel commented · Jun 21, 2023 at 06:55 PM

I do understand that MPTTs are over sized and will not produce full power. Even if you have 14 MPPTs with 50A fusing that is total of 700A parallel fuses. that to be added battery fusing will most likely easily exceed 1000A system rating. That not counting that inverters fuses. That's my understanding how that caution text in manual should be interpreted.

1 ·

screenshot-2023-06-21-at-214243.png (68.7 KiB)

cookinwitdiesel janieronen commented · Jun 21, 2023 at 07:48 PM

That "main system" fuse is essentially to protect the busbars themselves. Everything is already individually fused to protect their conductors. In this case that is:

2awg to the MPPTs (with 100a fuses)

Dual 2/0 to the Inverters (with 150a fuses on each)

4/0 to the batteries (with 400a fuses)

That all still has nothing to do with the fact that at 700a across the shunt/main fuse the thing is cooking itself to death and causing a fire hazard. Which is the main topic of this post.

System diagram is attached below. We ended up adding an inverter bypass between the input and output Breaker Panels as well as doubling the battery count to 20x with them in 10 parallel pairs wired to the lynx as shown below with 4/0.

Again, 48v would have been preferred but at the time there were no UL listed 48v options from Victron (there is now, since march 2023).

0 ·

ranch-solar-diagram-8-29-2022.jpg (1.9 MiB)

janieronen cookinwitdiesel commented · Jun 21, 2023 at 09:59 PM

Thank you for the system diagram. What ever you plan to do next you must change fusing strategy. Linked ANL fuse will trip way higher current than 1000A. 750A model trip for RMS Amps is something between 1000A-2000A which is way above Lynx specs.

Second reason is massive LFP battery bank short circuit current. ANL fuse is okay only for 6kA which is okay maybe for one of these beast of batteries. Victron does not provide fault current/short circuit values but instead battery internal resistance which is 1,5mOhm per one of your battery. That means batteries short circuit current is only limited cables and connections its that high inrush current capacity.

In example class-T fuses have better curve, can dissipate more heat and will safely stop arc. Main fuse reason is not only to ensure busbars & shunt safety but systemic maximum combined power in case of malfunction. In this case looks like main fuse is a reason the system heats up drastically.

As stated earlier it is highly recommend to build system with massive amount of parallel batteries, MPPTs and Inverters from the ground up with extra carefully. Tiny Fuse is under situation where it could be possible transferred as plasma in case of short circuit in any of system dozen connection points.

Lynx is great value for simple systems with few batteries, max three inverters and few MPTTs. It can withstand high peak loads but that many components are making system vulnerable for errors. Every each of additional connection between short busbars can lead dangerous situation if fails. All weak points should be minimised in system design and make as simple as possible. Quickly counting around 15-20 mechanical connection between battery to inverter in series. That is a lot of heat generated in additional to voltage loss on 37 fuses & shunt.

Lastly is great to see that size of system and hope you can build it even more awesome with large raw copper busbars and fusing that can keep you and your installation safe for years to come.

0 ·

screenshot-2023-06-22-at-02457.png (80.3 KiB)

cookinwitdiesel janieronen commented · Jun 21, 2023 at 11:00 PM

The system is already built and has been online since September 2022. Understanding the points you made, my main concern currently is that in normal operation (nothing failing or shorting) the ANL fuse currently in use is apparently getting dangerously hot while well within its supposed operating parameters. And that, combined with your points above, raise the question of why Victron would position the Lynx system for 1000a.

0 ·

janieronen cookinwitdiesel commented · Jun 22, 2023 at 02:53 PM

Something like Littlefuse JLLN800 would fit easily and there 1000Amp versions as well from other brands which can be fitted.

https://www.littelfuse.com/products/fuses/industrial-power-fuses/class-t-fuses/jlln/jlln800.aspx

1 ·

cookinwitdiesel janieronen commented · Jun 22, 2023 at 04:01 PM

The size is similar but not identical, will have to chew on this some more. Definitely seems like a better fit for the application.

0 ·

cookinwitdiesel cookinwitdiesel commented · Jun 22, 2023 at 04:04 PM

That is also a $400-$500 fuse, so I will be reluctant to drill/modify it any haha

1 ·

janieronen cookinwitdiesel commented · Jun 22, 2023 at 01:59 PM

Would it be possible to source class-T fuse 800-1000A rating? Many different models can be fitted and if cannot find with correct hole spacing there is possibility to drill new hole. Make sure cartridge is fitting between busbars so no need to modify lynx. These tend to warm a lot less and can protect your equipment.

0 ·

Answer 2 · 2023-06-22T12:46:38Z

pwfarnell answered · Jun 22, 2023 at 12:46 PM

Is the fuse being run too close to its rating. I can not find the tech specs for an ANL 750A fuse, but I have seen fuses are typically 100mV voltage drop at rated current. As a thought experiment, lets say that dV is 93mV at 700A, this means that the fuse would be generating 65W of heat which is a fair load to dissipate. Sure when a fuse blows it creates a very high heat load but that is short lived so there is limited time to heat the rest of the system. Running the fuse close to rated will generate the heat continuously, which will then soak into the system. Perhaps a 1000A fuse would be better, with a lower resistance and lower heat production for continuous running at 700A. There are various reference stating that the max continuous load is ideally 75% of the fuse rating or the fuse is rated for 125% of the maximum continuous load.

1

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

cookinwitdiesel commented · Jun 22, 2023 at 02:25 PM

I would love a larger fuse with lower Vdrop but it seems that 750/800 is the highest rating you can find on reputable sourced fuses for ANN/ANL form factor.

0 ·

Answer 3 · 2023-06-22T19:49:45Z

Johannes Boonstra (Victron Energy Staff) answered · Jun 22, 2023 at 07:49 PM

@cookinwitdiesel it has been discussed already as I see, but would like to add I do agree with most here that the fuse is used too close to its rating which causes heat generation when used on high power for longer periods. getting a higher rating would tackle this I feel. we have tested the system on 1000A , with large cabling 9which also divert heat) and 1000A fuse which did of course get warm , but did not had a burnt/deformed enclosure as result.

8

2 |3000

Attachments: Up to 8 attachments (including images) can be used with a maximum of 190.8 MiB each and 286.6 MiB total.

cookinwitdiesel commented · Jun 22, 2023 at 08:14 PM

Thank you for the input Johannes, can you share what fuse you used that fits and is rated at 1000a? I have not been able to locate anything on my own so far. Nothing else in the system is getting hot, just the fuse and attached bus bars so I am very confident that the fuse is the main source of heat here.

0 ·

cookinwitdiesel cookinwitdiesel commented · Jul 01, 2023 at 03:05 PM

@Johannes Boonstra (Victron Energy Staff) Can you share what fuse was used for this testing that fit in the Lynx Shunt? Thanks!

0 ·

Alexandra ♦ cookinwitdiesel commented · Jul 01, 2023 at 03:28 PM

@cookinwitdiesel

It is designed for the CNN fuse. Bigger ones come from littelfuse that can go up to 2500A.

In the manual.

0 ·

cookinwitdiesel Alexandra ♦ commented · Jul 03, 2023 at 04:45 PM

@Alexandra where are you seeing this larger sized fuses? I could only find mention of up to 800a

0 ·

Alexandra ♦ cookinwitdiesel commented · Jul 03, 2023 at 05:29 PM

I got (rather stupidly) messed up by the up to 2500A interrupting speel.... you are right they only go up to 800A. I apologise for not reading more thoroughly and leading you to wasting your time looking for one.

The CNN is a much slower blow fuse than the ANL which tends to blow right on its rating.

0 ·

screenshot-20230703-192021.jpg (106.5 KiB)

Alexandra ♦ Alexandra ♦ commented · Jul 03, 2023 at 05:36 PM

@cookinwitdiesel Further thought on your system, 1000A is alot to go wrong (as you know)

I would bridge the fuse with a nice fat bus bar of appropriate size; interrupt the Lynx system with NH fuse holder and 1000A NH fuses.

(Head not in the clouds this time) The NH is what we normally use on bigger amp systems, sinply for arc quenching and heat is not a thing to deal with.

Class T fuses and holders are a bit more elegant

1 ·

Alexandra ♦ Alexandra ♦ commented · Jul 03, 2023 at 05:52 PM

If all your batteries and mppts are individually fused to the bus bar, you could also just bridge it with a piece of bus bar and still be safe.

0 ·

cookinwitdiesel Alexandra ♦ commented · Jul 28, 2023 at 04:29 PM

EVERYTHING attached to the busbars is fused. The fuse at the Lynx Shunt should be to protect the busbar itself as a conductor I would imagine since collectively, this system can produce over 1000a (but never should as configured/programmed) it is appropriate to have an overcurrent protection on that conductor (the busbar itself).

0 ·

question