Schneider MPPT 60 150 Supercapacitor input

wildlaurel

I am hoping to use the Conext MPPT controller to fully utilize a supercapacitor bank (charged externally to 80v) to keep another 24v supercapacitor bank afloat at its output. This setup would power a 24v operated UPS for about 3 minutes until generator starts and takes over.

So shoutout to the experienced people on the forum: Would the MPPT controller act as a dc-dc buck converter in this scenario?

MPPT controller input (PV side):               supercapacitor bank 41F / 80v x 2
MPPT controller output (battery side):       supercapacitor bank 108F / 32v

Chain:

supercap 80v   >>>   MPPT cont.   >>>   supercap 27.6v   >>>   UPS   >>>   AC power

PS. Why the need for this weird setup? I am getting these supercaps of 41F/81v for $50 each and unable to source a dc-dc converter which can efficiently work from wide input of 80v down to 24v.

boB

Interesting application.

Yes, MPPT controllers are almost always buck converters and I would think that the hardware at least should be able to keep the output voltage constant.  

The only issue I see off the bat is starting the charging of the supercap from a discharged (low low voltage or even zero volts) condition.

The controllers typically need to start running from some minimum battery side voltage...  Maybe 10 volts ?

Dave Angelini

As long as 32V (out) is the max and only voltage required. One would just program bulk, abs, & float for 32V.

mike95490

I would worry about the MPPT circuit trying to find the max MPPT point of the 80V super cap bank, because the MPPT is going to be looking for a PV panel and bouncing the voltage around to verify the actual MPPT point.   Will the 80V bank ALWAYS be 80V, or will it's voltage sag when power is lost, as it pumps charge into the 28V cap bank of the inverter.  If it's voltage sags, then you can't use a fixed MPPT setpoint.

wildlaurel

boB said:

Interesting application.

Yes, MPPT controllers are almost always buck converters and I would think that the hardware at least should be able to keep the output voltage constant.  

The only issue I see off the bat is starting the charging of the supercap from a discharged (low low voltage or even zero volts) condition.

The controllers typically need to start running from some minimum battery side voltage...  Maybe 10 volts ?

I would charge it externally as the controller may not even detect discharged bank.

Dave Angelini said:

As long as 32V (out) is the max and only voltage required. One would just program bulk, abs, & float for 32V.

I think I will try 24v so that the bank at the input can discharge as much as possible.

mike95490 said:

I would worry about the MPPT circuit trying to find the max MPPT point of the 80V super cap bank, because the MPPT is going to be looking for a PV panel and bouncing the voltage around to verify the actual MPPT point.   Will the 80V bank ALWAYS be 80V, or will it's voltage sag when power is lost, as it pumps charge into the 28V cap bank of the inverter.  If it's voltage sags, then you can't use a fixed MPPT setpoint.

I have the same concern as to how MPPT algo would behave. I was hoping this following operating mode could be the key.
Yes, 80v bank would start sagging down until input matches output voltage and controller can no longer keep up.

Dave Angelini

I doubt it will be a problem for the mppt with a fixed voltage.
 If it was, you can turn off the auto track algorithm in the mppt and set it to a fixed voltage. We did this back in 2006 here when I did an  offgrid beta for Xantrex.

wildlaurel

Dave Angelini said:

I doubt it will be a problem for the mppt with a fixed voltage.
 If it was, you can turn off the auto track algorithm in the mppt and set it to a fixed voltage. We did this back in 2006 here when I did an  offgrid beta for Xantrex.

The supercap bank will be externally charged and floating at 80v but as soon as there is a utility outage, the bank will be loaded (by controller) and voltage will drop down rapidly as it discharges. Do you think controller would keep up in "manual mode" and what voltage should i set it to?

Dave Angelini

You are going to have to test it. I would not guess on this without alot more data on capacity. As I said, I do not think you will need manual mode. You will pick a battery range (24vdc) and set the setpoints to what you think the system will run at longest until the gen starts.

There is also the other way to do this that most people end up doing. Start the genset on utility drop and back-up the critical load with a computer UPS. Or, do what we do in California with critical loads that can't go dark during the many outages. Back the whole facility up and start the genset when you are down to 20% state of charge. The back-up is about 8ms with Schneider electronics.

I could not do any of this and went offgrid   Good Luck!

BB.

You have the controller designer "boB from Midnite" (and formerly Outback) saying that it should work...

I do worry that a quickly collapsing Super Cap input voltage could confuse some (especially older?) MPPT type charge cotnrollers (that would "measure" Pmp=Vmp*Imp every few minutes--And basically with each "new Vmp" value, the controller adjusts Imp to hold Vmp). A quickly collapsing (few minutes from 80 volts to 30 volts), I wonder if the MPPT controller will be "power limited" if it's estimate of Vmp cannot keep up (i.e. 80Vmp*10amps=800 Watts; 60Vmp*10amps=600 Watts, using "old value of Vmp").

I do not know about the details of these controllers (especially the newer ones) manage the Pmp=Vmp*Imp equation.

Perhaps setting Vmp manual at (30 volts, for example) would work. That should not limit the power output as the source will always supply enough current to keep voltage above 30Vmp "floor".

The cost of $100 for a pair of Super Caps certainly looks tempting vs 4x 6v @ 200 AH golf cart batteries (4x$100=$400 for cheap FLA type). Checking electrolyte monthly (and other maintenance) for ~5 year life(?).

A higher end MPPT controller runs about $600 or so (of course there are less expensive brands too). Vs "cheap" G.C. batteries + inexpensive AC float controller... No second bank of capacitors, etc.

Have you looked at the life of the super caps you are looking at? From what I have read, they might have an approximate life of 5 years too. Will you be able to get replacements (at that price) 5 years down the road?

A pair of Fullriver 12 volt @ 55 AH batteries in series (AGM--No electrolyte maintenance) should be able to output 50 amps for 30 minutes...

Which brings up the 3 minute start time for your genset.... What happens if the genset does not start--Do you need a graceful shutdown time of 10 minutes (to write data to "disk", etc.)? At least you have time, after a failed start, to signal the computer(s) to perform the shutdown with a pair of AGM batteries for $400 over the next 5-7 years (expected life).

https://www.solar-electric.com/fullriver-dc55-12-agm-sealed-battery-12v-55ah.html
https://www.solar-electric.com/lib/wind-sun/Fullriver_battery_DC55-12_specs.pdf

-Bill

wildlaurel

Dave Angelini said:

You are going to have to test it. I would not guess on this without alot more data on capacity. As I said, I do not think you will need manual mode. You will pick a battery range (24vdc) and set the setpoints to what you think the system will run at longest until the gen starts.

There is also the other way to do this that most people end up doing. Start the genset on utility drop and back-up the critical load with a computer UPS. Or, do what we do in California with critical loads that can't go dark during the many outages. Back the whole facility up and start the genset when you are down to 20% state of charge. The back-up is about 8ms with Schneider electronics.

I could not do any of this and went offgrid   Good Luck!

I will try to set up a test rig.

I intend to the same as you are suggesting.. APC UPS with supercap will act as buffer for 90 seconds until generator starts/warms up.

supercap 80v   >>>   MPPT cont.   >>>   supercap 27.6v   >>>   APC UPS   >>>   AC power

Also, I noticed that MPPT algo might load the PV side to its short circuit current so that it can determine MPP and Isc.

wildlaurel

BB. said:

You have the controller designer "boB from Midnite" (and formerly Outback) saying that it should work...

I do worry that a quickly collapsing Super Cap input voltage could confuse some (especially older?) MPPT type charge cotnrollers (that would "measure" Pmp=Vmp*Imp every few minutes--And basically with each "new Vmp" value, the controller adjusts Imp to hold Vmp). A quickly collapsing (few minutes from 80 volts to 30 volts), I wonder if the MPPT controller will be "power limited" if it's estimate of Vmp cannot keep up (i.e. 80Vmp*10amps=800 Watts; 60Vmp*10amps=600 Watts, using "old value of Vmp").

I do not know about the details of these controllers (especially the newer ones) manage the Pmp=Vmp*Imp equation.

Perhaps setting Vmp manual at (30 volts, for example) would work. That should not limit the power output as the source will always supply enough current to keep voltage above 30Vmp "floor".

The cost of $100 for a pair of Super Caps certainly looks tempting vs 4x 6v @ 200 AH golf cart batteries (4x$100=$400 for cheap FLA type). Checking electrolyte monthly (and other maintenance) for ~5 year life(?).

A higher end MPPT controller runs about $600 or so (of course there are less expensive brands too). Vs "cheap" G.C. batteries + inexpensive AC float controller... No second bank of capacitors, etc.

Have you looked at the life of the super caps you are looking at? From what I have read, they might have an approximate life of 5 years too. Will you be able to get replacements (at that price) 5 years down the road?

A pair of Fullriver 12 volt @ 55 AH batteries in series (AGM--No electrolyte maintenance) should be able to output 50 amps for 30 minutes...

Which brings up the 3 minute start time for your genset.... What happens if the genset does not start--Do you need a graceful shutdown time of 10 minutes (to write data to "disk", etc.)? At least you have time, after a failed start, to signal the computer(s) to perform the shutdown with a pair of AGM batteries for $400 over the next 5-7 years (expected life).

https://www.solar-electric.com/fullriver-dc55-12-agm-sealed-battery-12v-55ah.html
https://www.solar-electric.com/lib/wind-sun/Fullriver_battery_DC55-12_specs.pdf

-Bill

MPPT algo is obviously proprietary so only way is to experiment (in safe limits)..

The point is to avoid the re-occurring cost of batteries since there are frequent outages and batteries cant keep up for long anyway. I have an auto-refuel natural gas genset paired with an AMF for start/stop/shift overs within 60-90 seconds.

Supercaps should be good for 500k to 1M cycles so I think they should outlast everything else in the system..

Since I have a spare Conext MPPT and its around >95% efficient in dc-dc conversion, I will try to give it a go. The findings might help others with similar use-case..

RCinFLA

I assume the supercap packs have their own balancers.  You don't want to use an MPPT controller as the MPP search algorithm will not likely be compatible with a supercap profile.  A simple buck converter of sufficient power capability and current limiting should work fine.  You will also likely need a low voltage shutdown on converter with a reverse current switch to prevent pushing current backwards through buck converter causing its switching MOSFET from getting hot from body diode conduction.

Solar Guppy

You will have to have the battery side charged for the controller to power up as the electronics are all powered off the battery by an internal step down converter. In my lab work I used power supplies all the time during development of the XW-MPPT. The charger controller house keeping supply is limited to about 75V, hence why it uses the battery, not array for energy source.

The tracking won't care if it is a battery, super cap or any other energy source as the unit will be in current limit in these configurations, which you can program on the controller using the user settings menu. You may want to set the current to a value that works best for your application.

Just an FYI, this was mostly my design effort working under contract with Xantrex in the 2005 time frame.

boB

You may or may not have an issue with tracking the input of a supercap since it does not have the exact characteristics of a photovoltaic (pv) solar panel ?

wildlaurel

I am yet to test as I need to put capacitors in series/parallel with balancer circuit.

I suspect that if MPPT controller tries to determine maximum available current by loading the PV side, then super caps with their virtually instantaneous infinite current capacity will damage the input side.

Solar Guppy

Once the unit is in current limit, there is no active tracking, but I would manually set the current limit to a low value for your initial value for testing, which can be done from the unit buttons itself.

I never ran with super caps, but there is no inherent issue I can think of that would say it would be an issue.

WebPower

Be mindful that the supercaps farad rating might not be accurate. The Kansas State University wind turbine research showed that the caps they bought weren't of the stated values. You might have to measure them to be sure. I know I'm a rank amateur here but my supercaps circuit is working pretty well for the IQ7's MPPT algorithm. It sucks them dry pretty quick but they're 2.7v so a lot less storage than those 80v. I used a power supply to find the IQ7's MMPT operating range for my circuit. I'm not bothering with balancing. Cheap and easy.