Overvoltage Protection

Question

I just destroyed my $1,500.00 wind turbine's battery charging module. I think I need a surge & overvoltage arrestor to add to my system to protect its replacement.

We have a small family farm in Northern Ontario Canada.

We are off-grid, producing power via wind & solar.  We average 50% cloud cover during the year, but have 95% clouds, only 5% sun gets to solar panels, often for weeks in winter, so we must rely on wind at these times.

Our wind turbine is a 3.5 kW turbine on 95' tall tower.  It produces 3 phase wild AC which shouldn't exceed 600 V ac.  Wind turbine power goes through a disconnect, then 10 A fuses to a full wave rectifier that changes it to varying voltage (600 V dc max.), and at varying amps, but typically 150 to 525 V dc. If batteries are fully charged, the excess power is automatically dumped to resistance heaters by the rectifier. The dc voltage output by the rectifier is fed to a Morningstar Tristar TS-MPPT-60-600V-48-DB that is a MPPT (Max. Power Point Tracking) battery charger to recharge our 24 battery bank (6 V dc batteries @ 500 Ah each), connected together in a 48 V dc nominal battery bank (8 batteries in series to creste a 48 V string, 3 strings in parallel to add capacity)

Morningstar Tristar TS-MPPT-60-600V-48-DB (60 A max. @ 48 V dc output, 600 vdc max. input) says the input voltage MUST NOT exceed 600 V dc at any time, not even a fraction of a second, or it will destroy itself. See:   https://www.morningstarcorp.com/products/tristar-mppt-600v/

I had the Morningstar Tristar TS-MPPT-60-600V-48-DB offline, the voltage was 500 to 525 V ac fed to rectifier due to moderate winds, and power was going automatically to the dump heaters, so I switched the Tristar on to recharge the batteries. There was a loud BANG. Later, I found small black pieces inside the Tristar from what exploded. I think there was a momentary surge at or above 600 V dc at the moment I switched it on, so the Tristar was destroyed (~ $ 1,500.00 loss).

I need something added to my system that is high speed, that will stop all surges or spikes in a controlled manner, and automatically resets itself when the voltage drops back below 600 V dc.

A friend suggested a "crowbar"?  Reading online, I have learned of a Zener diode, and/or a MOSFET?

We need every electron we generate for powering our farm, and can't survive running out to wind turbine 30x per day to manually reset the system, change fuses, etc., so hopefully there is a passive system that consumes no power until it has to step in to save the day, then automatically cuts itself out of the curcuit as quickly as possible so all the power can go automatically back to charging the batteries.

What circuit do you suggest?

Addendum

I may have distracted everybody by too much info. The full wave rectifier and the MMPT battery charger are two off-the-shelf units that need to play well together. I designed the system, selected the various components, installed them, operate them every day, and maintain them. Opperational data exists for about a year, but is not the issue. The full wave rectifier has a switch to set max. output voltage to 500 Vdc or 600 vdc, and I have it set at 500 vdc. In spite of this, it appears it doesn't react fast enough to keep the voltage beliw 600 vdc. The MPPT unit is highly sensitive to overvoltage, and for that reason I installed a digital AC voĺts, amps, Hz meter between two phases on wind turbine output, rectifier input, so that I could see what was going on, manually protecting the MPPT for 600+ vdc input which will destroy it. I was watching the digital meter, which was showing 500 to 525 vac, so I decided to switch on the MPPT. I assume when the MPPT power source switch is closed, the MPPT's complicated electronics start to assess the new power source (max. voltage at min. current), and if all is OK, the MPPT will start to use it. As MPPT starts to draw power from the rectifier, and ultimately from the wind turbine, the voltage drops as the amps go up from zero. The rectifier sees the load drawn by the MPPT, the power produced by the wind turbine, and starts cutting back on the dump load heaters so as to balance the overall system. Meanwhile the MPPT starts to sense the optimum volt-amps so as to produce maximum power by the wind turbine. The power provided by the wind turbine is then used by the MPPT to operate a 5 stage battery charger so that batteries are not overcharged.

One possible solution is a "crowbar" circuit that, on detecting an overvoltage, totally shorts out the entire circuit, and requires being manually reset. Not practical for my situation.

I understand that a avalanche power MOSFET can react in nano-seconds, and modulate its conductance (ie. Max. 30 A at 600 volts= 18,000 watts max., $18.00 each) which should more than control a 3,500 watt wind turbine. What I need is a circuit diagram known to work, or likely to work, as I understand there are dynamics, surges, ringing, waveform, ripple, etc. issues that must be considered.

Hope that helps, rather than confuse you even more.

Answer 1

No manual switching is good unless just to switch off for safety. Power surges to charging a battery bank with mismatched voltages is like a short circuit and if the charger was smarter would sense battery voltage before driving current.
You need to keep the MPPT regulator on all the time and regulate the load dump heaters on the input from the turbine so the output voltage does not exceed 550Vpk

This can be done with 3-phase controlled Triacs before the rectifier bridge using a voltage regulator for a heater temperature , but instead sensing peak voltage. With internal filtering the input voltages peak can be 41% higher than RMS.

One commercial solution is called a Solar diverter. https://www.gregorheating.co.uk/wp-content/uploads/2020/01/eddi-datasheet.pdf

You must have speed limiters on the rotor blades as well for mechanical safety and also since voltage is proportional to speed if you don't have enough loads.

Investigate the battery charge settings and operation for power sequencing and observe what limitations exists in the manual.

A transfer switch that goes open in a 2way switch momentarily rise in excess voltage which is your problem, not a lack of over voltage protect.

This ability you have only needs to sequence the charger before you switch off the loads. This can be done automatically using a regulator to control the heaters that starts at 550V and shuts offjust below that.

Answer 2

If batteries are fully charged, the excess power is automatically dumped to resistance heaters by the rectifier

We'd need to know way more how that is configured etc.

The way the system is supposed to work is that the heaters are never off unless the charger has been consuming power for at least 5 seconds (for example).

Thus, getting battery charger turned on should keep the heaters going for at least another few seconds.

If, for whatever reason, the heaters are off while the charger is off as well, the heaters must be turned on about 5 seconds before the charger is turned on, and stay turned on for another 5 seconds.

I'm sure the control signals are available to synchronize this "dance" using a simple PLC to set up the timers and logic. In a pinch it could be Arduino with a relay shield, but there are excellent affordable PLCs for a few hundred bucks, designed to survive industrial environments.