solar diversion on line side of charge controller

[cardamon]

I have been searching through old threads for several hours and cant find much on this.  I have a (potential) situation where I will have about 3kw of Pv at 550 vdc connected to a xantrex mppt 600v charge controller.  For various reasons, the design will be such that I will have more PV than the charge controller can handle so I would like to line side divert using the same techniques as discussed in the "Water Heating Controlled by a Classic 150" thread started by Chris O.  That thread touched on this  toward the end and a few mentioned that it could be bad for the charge controller.  I am not a power electronics engineer, but I am not seeing the potential harm to the CC.  Its what Baleen whales do: just extract a little resource as they swim by.......Has anybody tried or know of any further information on this?  I think Chris said he was going to try it out.......

[SparWeb]

While mixing solar PV, charge controllers, and diversion loads is legit and useful in many cases, operating a CC at or very close to its Voc limit is asking for trouble.  Not just the CC, but the wiring you've connected it up with may also be rated 600V maximum, plus other switchgear on the line if you have a disconnect, fuses, etc.

What will happen on a very cold day (last week it was -30 here) to the Voc of the panels?

[cardamon]

Ok, several more hours of digging and I stumbled across dgd's post about it.  Looks doable.  They dont give away 1000VDC SSR though.....

Sparweb:   The 550vdc figure is voc at -22.  VMP would be a "child's play" 388

[joestue]

is 550 maximum or the nominal voltage?

remember, solar has a "knee" in the voltage/ amperage curve.

if you're dealing with 550 volts nominal you could have 800 volts open circuit, which requires 1200 volt igbt's.

if you have 550 volt maximum it is no problem to throw some 600 volt transistors at it and make a shunt regulator, and as such keep the voltage below 550vdc regardless of the amount of power available.

[cardamon]

VOC at -22 is about 550, VMP is 388.  I think the only thing I am not quite sure of is how bleeding off some dc between the array and CC will effect harvesting at max power point.  I assume I would size the diversion elements as close as possible to the voltage and current max power of the array (which will change a bit with temp of course).  In theory, this would be 388 volts at 8 amps, which would be a 48.5 ohm resister.  I could tweak that figure a bit if I wanted to calculate in the real world conditions of voltage drop and ambient temeperature, but 4 240V, 4500 watts elements in series will be real close to that.  Am I thinking about this correctly?  Basically in my mind it boils down to having two loads in parallel off an array, one being MPPT and one not.  In this situation, Will the MPPT load, be able to make the whole setup operate at max power even if the second load is not?

[joestue]

the mppt converter should self limit at its maximum current limit, and that will manifest as the pv line voltage rising above the mppt point.

I dislike pwm for resistive load management so what i would suggest is using say 8 or more resistors, each wired to an IGBT or whatever and turns on at a set voltage, each progressively higher.

provided the loads turn on well above mp voltage, the mppt converter won't even know they are there.
but if the resistors aren't sized correctly then you get a sort of negative resistance oscillator.

also what is the mppt converter doing? if its charging batteries then the load will of course be variable and not constant.
if you want to extract that last 20% of available Kilowatt hours out of the system then you will need to set up a more complicated load management system

[Mary B]

Also plan in a failsafe for when a heating element goes open

[cardamon]

Joestue,

Here is the overall picture.  There will be the 3kw array hooked up to the xantrex 600 mppt CC as discussed (it is a 1400 foot wire run hence this setup).  There will also be a 500 watt PV array, and 1 kw of wind.  24V battery bank.  I will divert any extra resource to water heating hence I would like to keep everything fully utilized and operating at MPP all the time. At 24V, the CC can only handle about 2250 watts, but I have a larger array for more input on cloudy days and because If I can divert off the line side of the CC, then I can utilize all 3KW despite the CC only being able to handle 2250.  There are of course a zillion different ways to set this up and I havent totally settled on the exact topology yet.  Of course I am considering any input from the forum.  My initial thought was to divert any extra from the 3kw array on the line side using a classic controlling a SSR and heating elements, the way Chris O is doing it but taking into account the line side potential inductive issues as dgd is doing it (http://midniteforum.com/index.php?topic=727.0).  Now that classic would also be connected to the wind turbine, and a separate diversion system using tristars set a little lower would regulate the wind and 500 watt PV array and perhaps "fine tune" any extra that gets through the 600V CC to 24V elements.  So basically, the 3kw array/600V CC and the wind part of the classic would run 'wide open' all the time.  One thing I am not quite clear on is will the 600v CC be able to operated the array at its max power point despite bleeding off some juice from the line side?  I am just having trouble grasping exactly how that works.

Your step method could work too.  Have maybe 4 elements that would total about 3kw and turn them on as needed.  Why exactly are you opposed to resistive PWM?  I have found that despite the noise it works wonderfully(currently though its set up "normal", no line side of CC diversion going on).

MaryAlana,  agree some redundancy is good.   However note that the diversion system discussed in the OP would be on the line side of the CC so if it failed the CC would regulate assuming I didnt disable or set it real high.......

[joestue]

I dislike pwm due to the current ripple it pulls from the line, and that ripple will be pulled out of your xantrex's input capacitors btw..

I suppose you could have the 600 volt line run in series through the water heating elements, use relays to short the elements out by default.
the panels will still be run at their MP point.

[cardamon]

why series and not parallel?

[boB]

I dislike pwm due to the current ripple it pulls from the line, and that ripple will be pulled out of your xantrex's input capacitors btw..

I suppose you could have the 600 volt line run in series through the water heating elements, use relays to short the elements out by default.
the panels will still be run at their MP point.

That's a possibility !  Normally shorted except when the batteries go to Absorb or Float or EQ (Voltage regulation stage).

Then, when you un-short part of that series resistor, the battery voltage will drop below set-point voltage and the controller will be forced into dropping its input voltage and go back to MPPT and of course heat up the dump load.

Not nearly as voltage controlled as the parallel dump load but this might have some merit.  Or, might be just fine ?

One thing, Caradaman about some of these controllers is that they are bidirectional in nature.  They create the very input voltage from the battery through the controller that they are making the PV source see at the controller's input terminals.  So, under certain conditions, the controller can have reverse current from the battery, through the controller and back out its input terminals into your dump load.   It's hard for it to drive the source though unless, for some reason the controller makes its input voltage go above the Voc of the PV array or higher than the rectifier rating of the wind or hydro circuit.  But a resistor dump load, it can certainly drive that way.  This would hopefully make the controller either disconnect or reduce its input voltage.

Also, a PWMing dump load will, as Joestue said, cause ripple on the controller's input capacitors and heat them up unnecessarily.  You can fix that by adding a diode with its cathode connected to the controllers' PV positive input terminal.  That takes care of the negative current from the battery and also help that ripple current in the caps.
It doesn't help the fact that the PV array is not putting out maximum power point voltage and power into that dump load though.  That is the job of the controller and its output (MPP tracked) goes into the battery side (the low voltage output).  The controller doesn't make the maximum power flow through to the dump load that is now on the input.

The system could do something about that though if there was a current sensor on the dump load and software to control the PV voltage so that actually happened but that is quite a different system.

It's kind of an interesting problem though I must say but there may be something I am not thinking of here.

boB

[joestue]

I dislike pwm due to the current ripple it pulls from the line, and that ripple will be pulled out of your xantrex's input capacitors btw..

I suppose you could have the 600 volt line run in series through the water heating elements, use relays to short the elements out by default.
the panels will still be run at their MP point.

That's a possibility !  Normally shorted except when the batteries go to Absorb or Float or EQ (Voltage regulation stage).

Then, when you un-short part of that series resistor, the battery voltage will drop below set-point voltage and the controller will be forced into dropping its input voltage and go back to MPPT and of course heat up the dump load.

resistors go in series with the 500 volt pv line, not the battery line.

Not nearly as voltage controlled as the parallel dump load but this might have some merit.  Or, might be just fine ?

putting the resistors in series preserves the mppt function if i'm not mistaken.
you'll still have to prevent oscillation with generous hysteresis in the relay controller.

One thing, Caradaman about some of these controllers is that they are bidirectional in nature.  They create the very input voltage from the battery through the controller that they are making the PV source see at the controller's input terminals.  So, under certain conditions, the controller can have reverse current from the battery, through the controller and back out its input terminals into your dump load.   It's hard for it to drive the source though unless, for some reason the controller makes its input voltage go above the Voc of the PV array or higher than the rectifier rating of the wind or hydro circuit.  But a resistor dump load, it can certainly drive that way.  This would hopefully make the controller either disconnect or reduce its input voltage.

Also, a PWMing dump load will, as Joestue said, cause ripple on the controller's input capacitors and heat them up unnecessarily.  You can fix that by adding a diode with its cathode connected to the controllers' PV positive input terminal.  That takes care of the negative current from the battery and also help that ripple current in the caps.
It doesn't help the fact that the PV array is not putting out maximum power point voltage and power into that dump load though.  That is the job of the controller and its output (MPP tracked) goes into the battery side (the low voltage output).  The controller doesn't make the maximum power flow through to the dump load that is now on the input.

two things there,
yes, the ripple current pulled by the pwm dump load could certainly be pulled out of the battery backwards through the converter.
this would wreak havoc on the mppt algorithm and possibly trigger complete loop control failure in the converter
could also lead to chaotic oscillation or shutdown, and there are hundreds of details that go into such a case.
the input capacitors should be big enough to stop that, but they may not be able to safely handle the additional ripple current.

but its easy enough to throw more capacitors on the dump load.. but the problem is no one seems to know why they need to be there...


if the mppt converter brought the pv line to ground, rather than letting it rise when the output current limit exceeded safe levels, then you would put the additional resistors in parallel across the input line.. again if i'm not mistaken.

its possible the mppt algorithm is preserved in both cases.

[hydrosun]

Before mppt was available I put in some hydro systems with heaters in series from the hydro to the battery. That allowed the hydro to work at a higher voltage than the battery and cut the amperage and power loss in the long wires from the hydro to the batery. The heat was needed anyway and the batteries were almost always full. So instead of losing voltage in the wires, it was lost in a heater where it did some good. I like this idea of a resistor(Heater) that can be bypassed if the power is needed to the battery and then using a mppt to keep the voltage high and wire losses low.
If the voltage is high enough then regular 240 volt heating elements in a water heater could be used. The thermostat would have to controll a fet or transistor that could handle the dc voltage. Or the thermostat could turn on the same bypass relay as the one controlled by the voltage trigger.
This might work on much lower voltage systems too. Instead of separate load diversion at battery voltage the diversion could be done at solar voltage minus battery voltage. This brings up new possibilities but the design has to be    carefully thought out.
Chris

[cardamon]

This is an interesting puzzle. 

One thing, Caradaman about some of these controllers is that they are bidirectional in nature.  They create the very input voltage from the battery through the controller that they are making the PV source see at the controller's input terminals.  So, under certain conditions, the controller can have reverse current from the battery, through the controller and back out its input terminals into your dump load.

  Very good to know, I was not aware of that.  I did know that solar converters made some bi-directional dc-dc converters but I assumed they sort of went out of their way to make them like that

Joestue, can you elaborate on your statement, I don't quite follow:

if the mppt converter brought the pv line to ground, rather than letting it rise when the output current limit exceeded safe levels, then you would put the additional resistors in parallel across the input line..

Are you saying that if the controller shorts vs open circuits the input to regulate?


Upon further thinking, there is yet another piece to the puzzle.  This system will have a very large amount of input relative to the battery bank size because i want to heat lots of water, and because i want to be able to use power tools but I am ok with being flexible with when I use them so that battery bank can be small.  This results in a potentially very high charge rate during bulk which would be nice to suppress. Those facts added to the original premise of wanting to utilize the power that is beyond the charge controller rating makes for some fun control schemes that would ideally cover all these scenarios and their permutations:

1.  Utilizing energy to heat water from the line side of the charge controller that is beyond its rating
2.  Limiting the charge rate to a max of say 40 rather than 80
3.  Allowing the CC to let thru all 80 amps when the inverter can use it (power tools, laundry)
4  Operate the panels at MPPT all the time.

It seems to me that regardless of a series or parallel connection the CC would operate the array at MPP as long as the load was in  the ballpark enough so it wouldnt make the CC have to work out of it mppt operating range.  This assumes that the cc is always doing something - if it was satisfied than of course the line side load would not run the array at mpp by itself.  Maybe it could be set up so that the CC line side diversion only does half to 3/4 the array energy max, and the battery side diversion does this rest and that would keep the CC "working" the array at MPP even when its all ultimately going to resistive elements?  That would also take care of the extra 800-1000 watts the CC cant process.  Now how to limit the battery charge current....Ok here it is (maybe!):  Use the CC  line side heating elements in series idea and size them such that they limit the CC max output to the desired charge rate (prob 40 amps) which will keep the batteries happy.  So I am eating up that extra 800-1000 that the CC cant and keeping the charge rate to spec.  Only problem now is if I fire up a power tool and want to run it off the sun, the CC is only letting the sun cover half the load.  So how about maybe a shunt in the inverter supply feeding a relay which would would short across the line side element thus taking it out of the circuit?  Problem though is need a way to also short out the element when the batteries do need juice and its a cloudy day - charging would be a priority over heating water.   That is something to think about.  This is fun.

[boB]

If you want to charge batteries and heat water after the batteries are charged, and do the heating at max power,
you are most likely going to have to either use battery voltage diversion for the water heating or have another
power point tracker circuit for the water heater if on the PV input side of the CC.  You can't really put the water heater in series with the charge controller input because when the batteries are full, you have to limit the current through that heater load if it is in series just so you don't over-voltage the batteries in Absorb or Float or EQ stages.

What people usually do is to place the diversion load across the battery so that they can still use the MPPT of
the CC.  The diversion load makes a nice shunt regulator in the battery voltage regulation modes (stages)
if the CC can control it that way.  (Like MidNite's Waste-Not and Outback's Opportunity load modes)

Charge controllers are normally buck converters and won't be able to short the input to ground.  They can only
bring the PV down to battery voltage.

boB

[cardamon]

boB,

My thinking was I would have both a battery diversion system and the array side diversion.  The CC would be set higher than the battery diversion system so it would always be working.  that way I figure it can keep the array at max power as long as I size that series resistor correctly and make sure not to drop the voltage the CC sees below its mppt working range.  The reason I am thinking of this rather convoluted system is its a long wire run and the 600v controller is expensive so just adding another controller and wire run is not going to happen.  Also, adding more panels is series is a free lunch in terms of wire losses so its just too tempting to keep going past the CC wattage rating up to its max voc rating!  Q:  Does a series resistor between the array and the CC give you the some potential inductance problems as a parallel resistor?

[joestue]

regarding my proposed series resistor idea:

without either simulating this system or having something similar to play with i'm not sure whether series or parallel resistors would be better but i think that series resistors would be much less effort to set up and get tracking properly.

but you have two issues here that are separate and have to be solved differently.
one is  that the max power is limited by the output current of that converter, the input could range from 600 to say as low as, 300 volts and that would have no impact on the maximum power it can push into the batteries, correct?

dropping the incoming line voltage with a series resistor provides a more aggressive mppt function in conjunction with the standard algorithms, than does siphoning off current with parallel resistors but they both achieve the same thing, moving the MP point just below ~80 amps into a ~26 volt battery.

either configuration can be set up to be controlled basically the same way.
both can get the panels to within 5% of MP.
 the difference is whether you use voltage setpoints with hysteresis, or current setpoints with hysteresis.
both are approximately the same work, however, the series resistors could theoretically use much cheaper switches, relays for example.

but the second problem of how to manage pushing 80 amps past the fully charged battery into another load,
i have no idea how to do that with off the shelf hardware, there may be some ways to do it.
does that mppt converter have remote battery voltage sense lines?

regarding inductance issues:

PWM of a resistive load consumes a great deal of current ripple, regardless if its a 10 ohm series resistor or a 500 ohm parallel resistor.
I don't propose either topology be PWM'd for your situation.
Both series and parallel configurations will have issues with resistor inductance but both are easily managed.

also, series resistors don't drop any power unless the converter is pulling current, so for that reason you may want to use standard water heating elements configured in parallel for that reason...but you will be forced to use IGBTs or solidstate dc relays or discrete mosfets to do the switching, and that will have to be done at up to 500 volts.

[cardamon]

JS,



 

but you have two issues here that are separate and have to be solved differently.
one is  that the max power is limited by the output current of that converter, the input could range from 600 to say as low as, 300 volts and that would have no impact on the maximum power it can push into the batteries, correct?

dropping the incoming line voltage with a series resistor provides a more aggressive mppt function in conjunction with the standard algorithms, than does siphoning off current with parallel resistors but they both achieve the same thing, moving the MP point just below ~80 amps into a ~26 volt battery.

either configuration can be set up to be controlled basically the same way.
both can get the panels to within 5% of MP.
 the difference is whether you use voltage setpoints with hysteresis, or current setpoints with hysteresis.
both are approximately the same work, however, the series resistors could theoretically use much cheaper switches, relays for example.

but the second problem of how to manage pushing 80 amps past the fully charged battery into another load,
i have no idea how to do that with off the shelf hardware, there may be some ways to do it.
does that mppt converter have remote battery voltage sense lines?

The more I think about it, I think the series resistor and staying away from pwm on the array circuit  idea is good.  One benefit is that it is sort of self regulating.  With a parallel resistor, I would need some control to take it out of the circuit during low output situations when heating water is not a priority.  With a series resistor, the energy dissipated by it at say 1/10 array output is negligible.  After crunching some ballpark numbers, it looks like I may need the option of two series resistors to account for coldest and hottest conditions:  The resistor needed to eat up that extra 1000 watts on cold days will bring the voltage below the 195V mppt range of the controller on hot days.  I was looking at a voltage controller relay that I think could handle that nicely.  It has a hysteresis function so it could switch to the other element if the voltage to the CC went below 200, and wouldnt switch back until it went back above 300 or whatever. 

Yes as mentioned above, I see the element(s) being sized so that the CC always sees voltage within its mppt range which is 195 to 550.  IDeally they will clip off the right amount of power so 79 amps still goes to the batteries.  I think the  battery diversion system would take care of both of these issues.  I can use some tri star diversion controllers which I already have to 24 volt elements.  The setpoint of these will be lower than the CC so as far as the CC knows the batteries are always hungry for all 80 amps.  Yes I believe that CC has an aux that can be programmed so that could connect battery side dump instead of the trisars as another option.

Assuming that all works as planned, the only other issue is if i want to limit the charge current during the bulk stage as my battery bank is very small for this amount of power, so that would require a separate battery side diversion that went off current rather than voltage.  Perhaps a dc current transducer could drive a pwm diversion controller.  Off the shelf dc current transducers always seem to put out 5-10v so I would need a doubler circuit then it could drive a 12v diversion controller.   I am not ass concerned about figurintg this issue out however as this system will rarely deeply cycle so it may not be worth worrying about.