Author Topic: Disaster - short and open circuit protection? (Read 12772 times)

Vortechs · « **on:** December 20, 2013, 06:09:24 AM »

Disaster struck my hydro last night, the rectifier failed and short circuited the battery bank and open circuited the generator...

Luckily I was lying in bed trying to go to sleep and heard the high pitched whine of the generator and ran out to it just as the rectifier burst into flames. I managed to pull out one of the wires and put out the fire but the generator is melted from the induced heat and probably scrap. How can I protect against this in the future, a fuse would protect against a short but would leave the gene open circuit? Thanks.

Flux · « **Reply #1 on:** December 20, 2013, 12:09:50 PM »

Always a difficult one for wind or hydro.

Fuses need to be about double the maximum rated current, they must never fail in normal use and particularly with hydro you need bigger than nominal rating as they age over time.With luck the fuse will then protect wiring and would most likely have operated in your case when a rectifier goes truly short circuit ( probably won't work for a single diode failure in a bridge)

More effective is a reverse current detector that will never operate in normal conditions but will operate with reverse current. Don't know what is commercially available, may have to build it.

Probably good to grossly oversize the rectifier to avoid failure.

Protecting the turbine when the protection clears the battery circuitr is another issue. Ideally hydro should be designed to stand 100% overspeed witthout damage. Far less easy with wind.

You may be able to arrange the loss of load to shut off water supply fairly quickly, depends on the type of turbine, a deflector on a pelton nozzle is excellent, most turbines can't be shut down instantly due to pressure rise in the supply pipe.

The other thing with protection schemes is that they are never normally used and when needed may already be dead, needs regular testing.. Watch dogs to watch the protection are often even more troublesome and fail when needed or cause shut down when nothing is wrong.

Simplicity and reliability and gereral over rating of components tends to work best in simple systems without constant monitoring .

Flux

hydrosun · « **Reply #2 on:** December 20, 2013, 03:51:57 PM »

I work on more than a dozen micro hydro systems and have had several rectifier failures. None caught on fire. But the heat has demagnified the magnets (Only one system with two turbines on a single wire) or scorched and made the stator wires brittle. All were fixable. What type of system do you have? I've mostly worked on harris systems or homemade with car alternators. Many I have replaced the original rectifiers with 1600 volt 100 amp rectifiers with better heatsinks. What is your head and pelton size (Rpm of turbine). Did the circuit breaker at the battery open during this event? Is your turbine running hot normally? Is the rectifier heat sinked well? On harris units I have fashioned a fan to slip on the shaft internally to keep th eunit cooler and it tends to slow it down and keep things cooler in overspeed conditions. I also drilled more holes in the case of some early harris units to provide more airflow.
hope this gives you some ideas you can use to make your system more robust.
chris

Vortechs · « **Reply #3 on:** December 20, 2013, 06:40:53 PM »

Thanks, guys. I have one of the first Smartdrive greenbox pelton hydros which was sold as a kit by Ecoinnovation in New Zealand: http://ecoinnovation.co.nz/c-31-hydro-parts-green-case.aspx It's been running non stop for almost 10 years. When I installed it I had 130 feet of head on a half 3 inch / half 2 inch pipe run , (I've recently upgraded it to all 3" pipe). The turbine has an 11" rotor, no idea what the rpm is now. It's a F&P Smartdrive charging a 24v battery bank, it was putting out about 15 amps.
The original rectifier was made up of 3 single phase rectifiers, it failed during a lightening storm a few years ago and just short circuited the generator causing it slow down to almost nothing. I replaced that with a 30 Amp 3 phase rectifier, and bolted it throught the aluminum case the same as the original one - I guess the pipe upgrade was too much for it. Are 3 single phase rectifiers better than one 3 phase one? I have a spare stator, rotor and another 30 amp rectifier but dont want this to happen again. I have no fuse or circuit breaker between the generator and the batteries, never have. The generator only ran open circuit for about 2 minutes, the plastic sheath around the soft iron cores have melted and oozed out between the windings. The windings test OK with a meter, no breaks. I'll see if it's still functional tomorrow.

hydrosun · « **Reply #4 on:** December 20, 2013, 09:04:52 PM »

You need a breaker on every wire coming off the battery. Your hydro has a limited amount of potential power but your battery has a huge potential. Your previous failure showed that the hydro doesn't self destruct if it is unloaded but the battery can send enough power to do damage. So a 30 amp breaker can't be tripped by a 15 amp hydro but will trip by a short circuit from the battery.
Three single phase rectifiers will dissapate heat a bit better than one three phase. But I use much higher capacity three phase rectifiers. The spikes of a alternator can be much higher than the battery voltage. And if you have a wire break the voltage can be much higher. I'm now putting in 100amp 1600 volt three phase rectifiers that have a better chance of surviving anaccident. A capacitor or voltage clamp will help to protect the rectifier and suppress radio noise. Your long wire to the hydro is an antenna to transmit and recieve. A lightning Mov may help protect from lightning induced voltage spikes.
Chris

birdhouse · « **Reply #5 on:** December 20, 2013, 09:05:53 PM »

chinese 150A rectifiers are pretty cheap. i use a similar model on my wind turbine and it has never had an issue. i did bolt a HUGE aluminum heatsink on with a very thin layer of heat conductive goo. thermal paste??

with 15A coming in 24/7 maybe even a small computer style muffin fan on the heatsink?? seems you could spare the juice to run it.

adam

http://www.ebay.com/itm/MDS150A-3-Phase-Diode-Bridge-Rectifier-150A-Amp-1600V-/300622780463?pt=LH_DefaultDomain_0&hash=item45fe83982f

SparWeb · « **Reply #6 on:** December 21, 2013, 12:42:58 AM »

Can't beat the advice on the circuit protection.
I can offer a suggestion or two on the rectifier, instead.
I had a similar short-circuit, several years ago. Fortunately for me it failed safe, with the short only shutting the wind turbine down, rather than blowing the batteries. Since then I've used a 3-phase bridge rated at 100 Amps with no trouble at all. I can't even feel it get warm, the voltage drop is so low. The 3-phase bridge was a used model, I would never have sprung for a new one. You might find one on e-bay. I am still grateful to FL forum member Dean "Drives" for straightening me out that day.

Some folks prefer to build a full-wave bridge using 6 individual stud diodes. Schottkys usually. Stud diodes are capable of dissipating a lot of heat when fastened through a large heat-sink. To give you a starting point, look up the "DO-5 diode" for the shape of it, and look at a NTE6094 to get an idea of the performance you want. I don't know if that NTE will actually suit you (especially on the voltage rating), but once you've had a look at it, consider what peak voltage and current range you really need, and adjust accordingly. Many major silicon manufacturers make a wide range of these, so it's a matter of selecting from the Digikey/Mouser catalog, when it comes right down to it. The heat sink must be large enough to do justice to the diodes. I can't think of a good source for a photo off-hand (one hasn't been posted on FL for a while, I think) to help illustrate my point. Basically start with a big fat hunk of aluminum or copper, and the more surface area it has, the better. There are design guides available around the net to help you do some calculations.

Vortechs · « **Reply #7 on:** December 21, 2013, 07:46:29 AM »

It looks like you're right, Chris, looking closely at the stator every 3rd pole is undamaged which suggests that 2 of the phase windings were shorted through the batteries causing the heat and maybe not the over speeding. I'll definitely install a circuit breaker, what's the best type for this job, and I take it the capacitor would be across the output of the rectifier - type and rating?
Thanks everyone for your suggestions, this is a great site, really appreciate the help.

hydrosun · « **Reply #8 on:** December 21, 2013, 12:10:09 PM »

Square D QO breakers are rated for up to 48v and will work fine on a 24v system. Higher voltage systems I use Outback breakers. I think bolting the rectifiers to the hydro with water going through the unit will keep them plenty cool. I've used 82 uf 150 volt capacitors on hydros before. Also Harris has used automotive alternator caps. The ones that have it's case used as the negative. The caps are put accross the dc output.
Chris

Frank S · « **Reply #9 on:** December 21, 2013, 01:02:49 PM »

I am wondering if you shouldn't have a reverse current diode and a magnetic contacter like an IE or a MEMA motor controller in series with your fuse or circuit breaker. That way as long as you had normal output from your turbine power would be going into the bank . If the turbine shorted out the breaker would trip or the fuse would blow. If the turbine stopped producing for any reason the diode would not allow current to flow in reverse this would open the contacts on the controller. The contacts need to be DC rated and if were me I would get one that is way over sized.
Personally when I build my Hydraulic power packs I use 1 or 2 sizes larger for my AC motor controllers and 3 sizes above the NEC recommended for the DC motors in the stand by pump motors I do this to lessen the possible midnight service calls.

Vortechs · « **Reply #10 on:** December 21, 2013, 02:44:50 PM »

Quote from: hydrosun on December 21, 2013, 12:10:09 PM

Square D QO breakers are rated for up to 48v and will work fine on a 24v system. Higher voltage systems I use Outback breakers. I think bolting the rectifiers to the hydro with water going through the unit will keep them plenty cool. I've used 82 uf 150 volt capacitors on hydros before. Also Harris has used automotive alternator caps. The ones that have it's case used as the negative. The caps are put accross the dc output.
Chris

A double pole like this OK: https://solarconduit.com/shop/circuit-breaker-qo230-30a-2-pole.html or is 1 or 2 singles better, (do both wires from batteries need a breaker or just one)?

I have my system back up and running on the smallest nozzle I have, the 30 amp rectifier is bolted to the aluminum body of the hydro case that has cold water constantly pouring over the back, it's just slightly warm to the touch so hopefully OK until I get hold of a higher rated one.

Thanks Frank, I'll have do do some research to get my head around that.

birdhouse · « **Reply #11 on:** December 21, 2013, 02:51:35 PM »

the breaker pictured in the link will accept two wires in and DIN rail out (2 connections via a disconnect ect) and yes, those are the breakers rated for DC use. someone earlier said they're rated for up to 48VDC, but my memory remembers they are only good for 24VDC.

if it's on the DC side of things, a single pole breaker should do fine, but a double pole is probably better.

any chance the water pouring over things is occasionally shorting the three phases of the rectifier?

just some thoughts.

adam

Mary B · « **Reply #12 on:** December 21, 2013, 03:27:47 PM »

48vdc I just looked them up. Thanks for the heads up I am rebuilding my mains panel and adding 8 more 6 volt 232 AH bats

Vortechs · « **Reply #13 on:** December 22, 2013, 06:06:05 AM »

That site says:

Quote

They are UL Listed for DC branch circuits up to 48VDC (not for use in 48V systems).

OperaHouse · « **Reply #14 on:** December 22, 2013, 08:22:55 AM »

This all seems very real to you now, but this is likely overblown. Diodes if used properly have a very high reliability.
A close look at the specifications might reveal there was a potential for failure. I use circuit breakers but they are
prone to problems at low voltages and can be inaccurate. Years ago a friend had some problems with breakers and ran dome
tests. The trip points were all over the place. With many modes of failure of your system, a breaker could stay engaged
when something went bad. Using diodes with sufficient safety margin should be reliable enough.

I think your best added protection would come from from a redundant diode at the battery side to prevent a back feed. Any
line going from a battery has the potential for fault. This would be added protection beyond the required current protection.

As I said before, fuses or breakers may not open in all failures. Redundancy of diodes may be a very practical solution.
Consider four diode bridges in parallel. A short length of wire to each diode bridge along with fuse on the input and output
will provide enough resistance for the bridges to load share. Fuse ratings would be a little less than half the value used
if there was a single bridge. This would insure the disconnection of the bridge in a failure and allow the system to continue
to operate.

Any lightening strikes lately? A few MOVs and inductor with local ground may be appropriate for that long line.

Vortechs · « **Reply #15 on:** December 23, 2013, 05:44:44 AM »

Thanks for your suggestion, I've bought a 150A 1600v rectifier to use but I'll keep it in mind for the future. We had a big lightening storm the night after my disaster when the hydro was off, they're not that common here, we maybe get one or 2 a year. I presume your talking about Metal Oxide Varistors? I dont know much about them, what kind of setup would you use?

XeonPony · « **Reply #16 on:** January 08, 2014, 07:17:22 AM »

one mov per conductor to ground then one move between each phase.

Vortechs · « **Reply #17 on:** January 25, 2014, 02:59:56 PM »

Are the Maxi fuses OK to use, and what size fuse should I be using? My maximum possible output is about 20 amps so would a 30 amp fuse be about right?

SparWeb · « **Reply #18 on:** January 25, 2014, 11:08:50 PM »

The problem with those is not really the fusing current rating or anything like that - it's the reliability in such an inexpensive part. Are you operating your system at 12V? I wouldn't use those at 24V or 48V either.

That said: I have a 24V system, and for a time I was using a bank of mini-fuses like those to protect the power supply wires of a datalogger and another auxiliary device. It depends on what you're using them for. On the line from the wind turbine, I don't think they can handle that.

Vortechs · « **Reply #19 on:** January 27, 2014, 02:59:07 PM »

That's my worry, they dont look reliable and I dont think I want my system to be dependent on them. What other fuses and holders are people using? How about the old Wylex re-wireable house fuseboxes, any good?

SparWeb · « **Reply #20 on:** January 29, 2014, 03:06:28 PM »

No way

You have to use DC-rated fuses to protect DC wires. One major misunderstanding (I've done this many times) is the believe that fuses must protect the device. They don't.

Fuses are supposed to protect wires from over-current and possibly catching fire. The chance that a fuse may also protect a device from self-destruction should be considered a side-benefit.

The way to think about it is that a device that has failed should not be allowed to propagate its failure to other devices or to the power supply wires. If that failure could cause a short or over-current fault that would cause its supply wires to melt/burn/arc, then fusing must be in plate to prevent that from happening. The device SHOULD HAVE been designed to fail-safe so that it doesn't self-destruct, but that's not always the case with hobby/DIY stuff, is it?

But if you take that philosophy to heart, even in your DIY RE system, then you should look at the thing as a chain of devices, each of which should not harm the other. So the maximum output power that the generator can produce, multiplied by a safety factor (an extra 50% I guess), determines the maximum current that can flow from generator to battery. If the wires are sized so that this current does not cause them damage, then you have one step in the chain taken care of. Now look at the rectifier. Is it sized to withstand that maximum current figure? Under all conditions? On your heat sink? If so then you have taken care of another step in the chain. Another step: the diodes have a number of failure modes, one of them is open-circuit and another is dead short (as you've seen). A circuit analysis can give you a ballpark idea of the current that flows in a short circuit but you don't have to bother, because it will be greater than the normal current, by a lot. This is where the fuse comes in. Select the fuse to allow the normal maximum current, multiplied by that safety factor, and then round up. That fuse will allow all normal operation, and prevent the wire from melting if the diodes blow.

Because I assumed that the diode could blow, and then the fuse blows next, I have purposefully NOT designed the fuse to protect the diode or the battery. It only protects the wire. And it also happens that the battery doesn't get drained flat or overheated, but remember, that's just an extra benefit of the fuse protection, not the purpose. By blowing the fuse now, the generator may overspeed because it will now run unloaded. Note that the diodes other failure mode is open-circuit, so it's the same difference. Designing the generator to not self-destruct under this condition is the last step in your chain of safety. It may already be able to handle that, in which case, you are done.

OperaHouse · « **Reply #21 on:** January 29, 2014, 04:00:48 PM »

I have a great distrust of fuses and fuse holders. If you need to protect the battery I prefer to use solething loke this FWP-50A
that you can make a direct connection to.

As I said before I like redundancy in diodes. And when it comes to diodes, you may not always e getting what you are paying for.
I needed to get some bridge rectifiers fast and selected a KBPC5010. These are supposed to be 50A 1000V. These were marked BEN
and the seller mentioned that some failures had been reported as low as 27A. I was replacing a single bridge in a 20A charger in
a friends high lift. There was space so I doubled up two bridges and used about 5 inches of wire on each terminal to add some
resistance so the bridges would load share. With everything assembled I plugged in the charger, the meter indicated 15A going
into the battery. At only about 8A into each bridge, I was expecting them to run pretty cool. They were scorching hot. Later
I tested these diodes on the bench with several other types of 25A bridges. All the KBPC5010 diodes had a forward voltage 25%
higher at a test current of only 2A. I'd show more concern for the diodes you are using.

Mary B · « **Reply #22 on:** January 30, 2014, 03:12:05 PM »

I am using ANL fuses, common in car stereo applications with big amplifiers, relatively inexpensive and fuse holders are easy to get on ebay. Rated 80 volts DC 35 to 750 amps. http://www.cooperindustries.com/content/public/en/bussmann/electrical/products/supplemental/cable_limiter/ann_anl.html

Vortechs · « **Reply #23 on:** January 30, 2014, 05:50:02 PM »

The only difference I can find in AC and DC fuses is DC fuses are wider because DC can arc farther and for longer, but is that relevant to 24V? The gap between the electrodes on my immersion is narrower than a standard AC house fuse, if it was going to arc then it would be jumping that, no?

birdhouse · « **Reply #24 on:** January 30, 2014, 07:12:11 PM »

+1 on the ANL fuses. i use them for all my "important" fusing.

Mary B · « **Reply #25 on:** January 31, 2014, 03:41:31 PM »

Positive and negative battery posts get anl fuses, and the inverters each have one. My smaller load under 10 amps for lighting is your standard car fuse, they are rated to 32 volts.

News:

Author Topic: Disaster - short and open circuit protection? (Read 12772 times)