Building a clipper

[bigrockcandymountain]

So I would really like to build a clipper type controller for my wind turbine to add some redundant safeties.  I already have a 100a 3 phase ssr.  I will power it off the aux2 of the classic 200 charge controller set on pvV high. I am thinking of adding a voltage monitoring relay. 

The GRV8-01 looks like the right fit.  The ac/dc 110/220v should work i think. 

Can i monitor voltage between 2 phases of the 3 phase? Or should i get a small rectifier and feed rectified dc into the monitoring terminals?

I will have to find a 12 or 24v source to feed into the relay/ssr.  I am thinking a small dc-dc step down. 

I would rely on the classic aux 2  for most of the control of the ssr.  The voltage monitoring relay would be the back up if the classic fails or shuts down and the voltage goes high. 
 
I looked at kitestrings schematic and it is similar. I would add a diode to each positive control wire on the ssr.

The 3 phase load will be water heater elements in a tank close by.  9.6 ohms hooked delta.  It works out to dumping 4500w at 120 line voltage.  That is more than the turbine will make, so it should keep under control around 90 or 100v line voltage.  Hopefully that is close.  I think it is.   

Am i missing anything? Is there a better way?

[ChrisOlson]

I already have a 100a 3 phase ssr.
[........]
I would rely on the classic aux 2  for most of the control of the ssr.  The voltage monitoring relay would be the back up if the classic fails or shuts down and the voltage goes high.

I assume you mean a three-pole SSR. SSR's don't have phases. You don't actually need a three-pole. A two-pole works fine because one phase of the generator can be hooked to the load all the time. It doesn't do anything until at least one of the other two phases is hooked to the load.

If you have a four-wire generator with a neutral wire hooked to the wye connection in the generator, then it will produce single-phase power from any phase and you would need a three-pole relay.

I have run one of these things for the past 7 years without using any sort of external voltage monitoring relay with a Classic 150. Even if the Classic over-volts the aux outputs still stay active within the over-volt range. It just removes the load from the turbine by stopping power output to the battery if it over-volts. And then the clipper must handle it all. The turbine runs most efficient when it over-volts the Classic and puts out the most power. Don't underestimate what you need for a clipper to absorb it. An 11 foot machine with a suitable generator can easily produce 4,500 watts running in over-volt range.

I would recommend using lower resistance elements, like 7.2 ohm. They will absorb more power than 9.6's, which are 6,000 watt 240VAC water heating elements. At 100VAC three-phase, three of those elements wired delta will only absorb 2,080 watts, and at 150VAC will pull 4,680 watts.

At 100VAC three 7.2's will pull 2,780 watts, and at 150VAC will pull 6,250 watts.

The load curve is a lot steeper with 7.2's to keep the turbine speed below 450 rpm, which is the structural limit for most 11-12 foot diameter rotors. At least the way wind turbine blades are made. If you have helicopter blades on it then you can run them at 560 fps blade tip speed with a 12 foot rotor, which is about 900 rpm. But helicopter blades are quite a bit more advanced construction than wind turbine blades are. The blades for an ultralight helicopter with a 12 foot rotor are about $12,000 apiece. The aluminum blades for a Bell Jet Ranger 206B, which run at 765 fps, cost $180,000 apiece and they are life-limited to 4,500 flight hours. Since wind turbines run cheap blades made of wood or hollow core composites the speed has to be kept down or they will explode. They cannot run anywhere close to blade tips approaching the speed of sound like helicopter main rotors and airplane propellers do 

[bigrockcandymountain]

Yep, you are correct.  3 pole ssr. 

That is very helpful to know the classic aux output remains active in hyper voc.  I will go ahead and just use the aux 2 without another backup for now.

I will try the 9.6 ohm elements just because i have them.  If i need to change them, i certainly will. 

I have nothing right now with my furling set very very loose. It has been working fine but with limited output.   It's a spring, not a weighted tail so infinitely adjustable.  Turbine is 13' rotor on a 6 pole 5hp induction motor conversion.  It's a heavy beast.

What is your aux output set up like?

[ChrisOlson]

I'm on my linux laptop where I don't have the MidNite Status Panel. When I reboot this thing to Windows (for whatever odd reason that may be) I'll start up the Status Panel and get the settings for the clipper. I know it starts clipping at 145VDC into the controller.

[ChrisOlson]

It looks like I have the width set at 2. I don't remember what BoB told me about how that works. I think it's a pwm pulse width over time, although the manual says something about voltage width for that setting. I only know at this point it's been set to 2 and been working that way for 7 years.

[kitestrings]

I would recommend using lower resistance elements, like 7.2 ohm. They will absorb more power than 9.6's, which are 6,000 watt 240VAC water heating elements. At 100VAC three-phase, three of those elements wired delta will only absorb 2,080 watts, and at 150VAC will pull 4,680 watts.

At 100VAC three 7.2's will pull 2,780 watts, and at 150VAC will pull 6,250 watts.

Chris, I follow your logic, but not your math.  I'm coming up with 3,125w & 7,031w on the 9.6 ohm resistors at 100VAC & 150 VAC respectively.  The 7.2 ohm resistors I'm getting 4,167 & 9,375w.  What am I missing?

Regarding the HyperVOC - read carefully what Chris is saying BRC.  The Classic is not destroyed in HyperVOC; it protects itself, but it is not active in the sense of managing the turbine loading in this range.  I believe any load into the batteries is gone in this mode.  It is solely left to the load bank (and furling). 

[bigrockcandymountain]

Perfect that gives me a start point.  Thanks chris.

My math also doesn't agree with yours chris.  I am also not very confident with 3 phase.  Maybe we are talking different voltage?  I am using line to line rms voltage. 

I understand the classic quits loading the turbine, but as long as the ssr is full on to the load bank, i should be ok.  What does yours behave like kitestrings? Does your voltage monitoring relay trip sometimes or only in case of a failure?

[kitestrings]

The voltage monitoring relay is intended as a fail-safe; should the Classic fail.  I have tested it, by intentionally lowering the set-point, but otherwise I don't think it has operated.

Our overall approach is a bit different - we have the load bank, homemade clipper, coming in at ~120-125V, IIRC, but then just slightly higher, 5 volts or so, is the actuator for the furling.  In high sustained winds, the furling actuator will shut things down.  This is much softer and I can also do it by preemptive choice - so if the remnants of hurricane Dorian look to be too rough, we'll just sit it out.  It saves a lot of wear and tear; and angst.

[ChrisOlson]

Chris, I follow your logic, but not your math.  I'm coming up with 3,125w & 7,031w on the 9.6 ohm resistors at 100VAC & 150 VAC respectively.  The 7.2 ohm resistors I'm getting 4,167 & 9,375w.  What am I missing?

In delta you have two resistances wired parallel. One phase is always inactive and not conducting. The formula for resistance of parallel circuits is 1/Rt = 1/R1 + 1/R2. For two 7.2 ohm resistors the calculated resistance is
1 / (2 * 0.139) =  3.6 ohm.

While I didn't go thru your math, you might be using three in parallel, which would be:
1 / (3 * 0.139) = 2.4 ohm

The reason you can't use a triple parallel circuit here is because of the way three-phase power conducts and the timing of the phases. Power is being conducted thru any two phases at a time.

Regarding the HyperVOC - read carefully what Chris is saying BRC.  The Classic is not destroyed in HyperVOC; it protects itself, but it is not active in the sense of managing the turbine loading in this range.  I believe any load into the batteries is gone in this mode.  It is solely left to the load bank (and furling).

This is correct. The Classic handles over-volt very well. Because I don't have a prop governor that varies the pitch of the blades I use furling in extreme wind to keep the speed below 450 with a 150V Classic. The machine is producing 185VDC open circuit. But with the "clamp" on the AC side with a three-phase load the over-volt is kept to <160. While the Classic isn't charging the battery at that speed and voltage, the AC clipper is running at full tilt keeping the machine under control (and producing tremendous amounts of power at quite high efficiency).

In over-volt I have 110VAC power @ 75Hz and max 22 amps continuous That is up to 2,400 watts on water heating continuous. It will produce more than that in really high winds, but the duty cycle of the generator is 25 amps for an hour if the outside temp is above about 30 deg F. If it's below freezing so we got good generator cooling it can do 2,500 watts continuous on AC load. But there is spikes where I've seen it send 40 amps to the water heaters before. Obviously, my original clipper could not take this. It got too hot. That's when I came up with why not use this for something useful, that is about 92% efficient electrically from the generator to the load? Water heating.

This can happen on any day when the battery gets to absorb too, even in 10-12 mph wind. The Classic unloads it as the solar takes over and it winds up into HyperVOC running at very high TSR. The AC load comes in and it starts heating water with three-phase power. We don't get no 2.0 kW at 10-12 mph. But it's still VERY efficient electrically, to the load.

It's been running this way for 7 years and I haven't been able to burn up a Classic controller yet. And I'm just using a 150.

When I first started messing with the Classic I tried various engine-driven generators in the shop. This is where I found out the Classic does not like "dirty" power input. It gets really hot if the DC power has ripple from a lot of THD on the AC side of the rectifiers. And the 20-pole high-frequency axial generator design was born. That was actually Les' (Flux) idea. He told me, if you want clean power you need higher AC frequency. Either gear up the 12 pole and run it fast, or put more poles in it. I tried both designs. While the geared design was successful, the 20-pole built to super-accurate specs puts out the cleanest power.

[ChrisOlson]

guys, I had honestly forgotten I do have a safety shutdown for the turbine. We don't look at our system much anymore. I put water in the batteries once a month. We refill the generator gas tank twice a year, and change oil in the generator twice a year. That's it. Our ComBox sends us an email once a week with our system summary, and I browse thru that but I don't actually fiddle with our power system anymore.

So I tend to forget some things.

It dawned on me that I had a RD-1 Relay Driver once and I sold it to TomW because I figured I didn't need it anymore. But then I did need it, so I bought another one. That RD-1 turns on the water heater element (powered by the inverter) when there is sufficient power available to heat water with the top element in our second pre-heater.

But I also have that RD-1 programmed to monitor the voltage on the SSR primary and the voltage from a current sensing relay on the water heater/clipper. If there is voltage on the SSR primary, but nothing from the current sensing relay it closes a mechanical contactor that shorts the AC side of the turbine and stops it.

I had forgotten I put that in as a safety to stop the turbine if the thermostats turn off the water heater, or one of the elements burns out. But I don't recall in the past 7 years that I've ever seen that come on. The turbine was heating water just a few minutes ago, I happened to see that RD-1 and I go, "ahhhhhh..... I wonder if that still works?" The output to the shutdown relay is on Channel 4 on the RD-1. I pulled the wire off the current sensing relay and the RD-1's Channel 4 green light came on and instantly shut the turbine down. It still works   

It doesn't protect against the Classic shutting down, which I suppose could happen. But I've never had that happen. At worst I'd maybe be out a set of blades on the turbine if it did. The tower and heavy build of the turbine will take losing one blade with no problem. It would likely shed the other two if it lost one due to the vibration and flexing that would cause the other two blades to hit the tower and blow those off too, which would then stop it.

[kitestrings]

Our ComBox sends us an email once a week with our system summary, and I browse thru that but I don't actually fiddle with our power system anymore.

I like that...we've got some similarities in set-ups - we preheat dhw also - but the only update I get is when I go home at the end of the day and it still is in the air.

Sorry to belabor this one, I get the series parallel relationship, but I had thought in a 3-ph balanced system that Wdel = 3(V^2)/R.  In delta Vph = Vline, so wouldn't the it be 3*(150V^2)/7.2 = 9,375w?  The only way I get 6,250 is in an open delta, with two resistors.  Is that what you have?  Maybe I missed this...

[bigrockcandymountain]

Well hey, i would consider that a great setup if you forget how it works.  That sure means it isn't causing you much trouble.  That is fantastic.

If i do add a voltage monitoring relay, i will probably have it activate an actuator on the manual furling.  That will be in the future. 

Thanks for all the ideas.  Now i just have to get on it.  Stuffing it all in the existing box will be the biggest challenge i think. 

[ChrisOlson]

Sorry to belabor this one, I get the series parallel relationship, but I had thought in a 3-ph balanced system that Wdel = 3(V^2)/R.  In delta Vph = Vline, so wouldn't the it be 3*(150V^2)/7.2 = 9,375w?  The only way I get 6,250 is in an open delta, with two resistors.  Is that what you have?  Maybe I missed this...

Ah, I see what you're getting at. You're calculating apparent power, which is what the load could draw being a purely resistive load with PF = 1.0.

What I'm calculating is what the load has to be to clamp the shaft power at PF = 0.81 to keep the rotor of a wind turbine from spinning faster than 450 rpm @ 4.6 kW shaft power with a generator dissipating 860 watts. It is a Y-Delta system and Vph is actually 65 volts using the real-world value.

So your load's theoretical dissipation in apparent power is one thing, vs what it HAS to dissipate at the design parameters of the generator is two different things.   

I know from experience the 9.6 ohm elements don't work. Tried that - they are common as dirt at Menards. They don't have a steep enough dissipation curve at what the generator actually puts out to control the machine. They might work for a less powerful turbine, that I don't know. But mine needs the 7.2's.

So, using the single phase parallel resistive calculation works for this, as I need to limit current per phase in a wye-wired generator. It has to operate at high enough voltage to not over-current a phase, while still being able control 4.6 kW of shaft power (combined with heat dissipated by the generator, which is also absorbing shaft power).

What it amounts to is a matched delta load for a wye source, with the wye source operating at its maximum capability, but not over it's maximum which would destroy it. The turbine is not powerful enough to actually power three 7.2 ohm resistive loads wired delta at the true or apparent power that load could draw. But the load curve of that load vs Vline matches the turbine.

I don't think MidNite's clipper is designed this way. I've never seen one, but I think it relies on pwm pulsing of the load, which is grossly oversized. When mine comes on it pretty goes to 100% pwm based on DC input voltage to the Classic, the load is matched to the turbine's power curve, while keeping the DC voltage inside the Classic 150's over-volt range.

I originally didn't even have a SSR on it. I had a mechanical contactor that just turned on. I put the SSR in later because the points in the mechanical contactor went bad.

[ChrisOlson]

kitestrings, you were right. It has been 7 years since I looked at this. I had to go in the basement and study my wiring on the water heaters. The first preheater is wind turbine only and with the shutdown relay off it is open delta. With the shutdown relay on it connects the top element on the second preheater in delta and stops the turbine gently.

The bottom heater element in the second preheater is hooked to a SSR in the power room, which is turned on by the RD-1 when there is sufficient power to run it off the inverter (it puts extra load on the battery bank/charging system to take advantage of extra solar during absorb as amps to the battery drops).

So I dug out my original chart of the power curve of the clipper. Hope this image handling thing works right here because it was a spreadsheet so I just took a screenshot of it:



The blue line was maximum continuous battery charging (which actually works out to about 2,500 watts to the battery). The green line is maximum continuous on clip. The red zone is rotor overspeed. But it never actually gets to the red zone (which is what we want on clip) because the power curve of the clipper load starts to increase faster than what the turbine rotor can produce, and that's what controls the speed of the rotor and prevents it from over-amping. The TSR drops and it loses power between the blue and and green lines. It never actually goes over 110 VAC (on the line).

The shutdown relay on always has enough load to stop it. But it doesn't just slam on the brakes like shorting it out. It's more of a gentle stop (kind of a necessary thing with a high-speed turbine). But that hasn't been used in 7 years (that I know of) until today when I tested it to see if it works yet.

I should rethink that shutdown relay. It relies on voltage to the SSR (which comes from the Classic). Plus current on the line to the clipper. If there is voltage, but no current, the RD-1 will stop the turbine. If there is no voltage to the SSR (Classic goes to hell) it never will stop it. That would give me something to actually do with my system for a change   

Trouble is, I don't know of a better way to do it. The idea, originally, was to stop the turbine if the SSR quits working.

[bigrockcandymountain]

Ok, so i bought a voltage monitoring relay.  I have a couple questions on how to hook it up. 

It is rated ac or dc.  Can i just hook it between 2 phase wires ac? Will the low frequency 120 degree 2 phase cause trouble? Would it be better to put a tiny 3 phase bridge rectifier on the input so it is dc in?

I will hook it in parallel with the classic aux 2 into the control of a 3 pole ssr.  I assume a diode on each positive wire would be a good idea?
I will get my 12v for the ssr control from the battery bank i think.  It won't ever draw unless the system over voltages.

Comments welcome

[MagnetJuice]

The rated supply frequency for that relay is from 45 Hz to 65 Hz. Would that be adequate for your application?

Ed

[bigrockcandymountain]

Ah i missed that rating Ed.  Thanks for that.  No it would be around 20hz or lower.