Charge Controller excess power... dump as heat or disconnect??

[joerossjr]

I am trying to understand exactly how a Charge Controller (CC) works when the batteries are full. Does it dump the excess power as heat, or disconnect output completely when the batteries are full? Is there a level where it turns back on to recharge the batteries? I have never seen this discussed before.

I ask because I am wondering what output I should expect to see. For a more sophisticated CC that does multi step charging (trickle, etc), I'd expect it to dump the heat since it has to have some output. But for a simpler CC, I would expect there to be zero output current.

Is my understanding correct? Is the more sophisticated simply a 300W heater when the batteries are full?

[sensij]

A PWM controller disconnects. An mppt controller adjusts the operating point of the PV system so that the irradiance is not as efficiently converted to electricity by the photovoltaic cells, so the "excess power" is never seen by the charge controller and is more or less distributed (and reflected+emitted) as heat over the array.

[PNPmacnab]

What many people don't think about is that getting an efficient charge controller often doesn't matter. By design most systems have over capacity and that excess potential power is just wasted. In a really smart system that excess energy can be used to heat water or operate appliances when there is a surplus. When the batteries are charged the excess energy is not dissipated, It is just wasted from doing anything useful. Smart systems can reduce battery size and are the future.

[AzRoute66]

There will always be extra waste as long as the charge controller refuses to recognize the difference between the battery and the working load. If they would just attach the working loads to 'proper' load terminals of the controller, then the controller could service, in order: 1. the working load, 2. the battery charging function, 3. excess production which could go to grid, compressed air, flywheel, hydro pumping, etc. storage. Note this is what utilities do with excess turbine production in a hydro system - they pump water back up to the top. Not very efficient, but better than running it through a superfluous heating element or other dump load. The argument coming back is that loads can be more than one charge controller can provide (especially when it is dark) so they must be attached directly to the battery - this is clearly untrue, there is nothing that says the wires from the battery to the controller are uni-directional.

Note that designing the charge controller in this fashion would allow for very precise battery SOC information as well. It might not replace the hyrdrometer completely, but would allow for a tool that falls somewhere between specific gravity and voltage. It would be especially valuable for AGM systems.

Such a [MPPT] controller could operate in a 'battery priority mode', that calculates the power available and lets the user (or user equipment) know how much power is available as 'excess' after load and battery service, making it possible for the user to know when blow drying their hair won't be impacting the battery charging process.

The whole concept is the elephant in the room that nobody will acknowledge and it is ridiculous.

In other words, it is well past time that the charge 'controller' start 'controlling' the battery charging process.

[SunEagle]

There will always be extra waste as long as the charge controller refuses to recognize the difference between the battery and the working load. If they would just attach the working loads to 'proper' load terminals of the controller, then the controller could service, in order: 1. the working load, 2. the battery charging function, 3. excess production which could go to grid, compressed air, flywheel, hydro pumping, etc. storage. Note this is what utilities do with excess turbine production in a hydro system - they pump water back up to the top. Not very efficient, but better than running it through a superfluous heating element or other dump load. The argument coming back is that loads can be more than one charge controller can provide (especially when it is dark) so they must be attached directly to the battery - this is clearly untrue, there is nothing that says the wires from the battery to the controller are uni-directional.

Note that designing the charge controller in this fashion would allow for very precise battery SOC information as well. It might not replace the hyrdrometer completely, but would allow for a tool that falls somewhere between specific gravity and voltage. It would be especially valuable for AGM systems.

Such a [MPPT] controller could operate in a 'battery priority mode', that calculates the power available and lets the user (or user equipment) know how much power is available as 'excess' after load and battery service, making it possible for the user to know when blow drying their hair won't be impacting the battery charging process.

The whole concept is the elephant in the room that nobody will acknowledge and it is ridiculous.

In other words, it is well past time that the charge 'controller' start 'controlling' the battery charging process.

IMO the intelligence of a charge controller is directly related to it's cost. You can get cheap PMW types on the internet or go with higher end MPPT type. The first is nothing more than an on/off switch the second can have much more functionality based on the manufacturer and price.

Still the problem with knowing a battery SOC requires more than just measuring the voltage and is also related to the type of chemistry used. All that type of measuring devices can be developed but at what cost and what market value. Maybe relying on battery SOC is not a priority for the equipment manufacturers.

[AzRoute66]

SunEagle , Good point. And look how much better it is when the charge controller exhibits more smarts (MPPT) on the panel side of the process. Everybody will agree that MPPT is a good thing. If the charge controller had more smarts on the battery side, people would soon come to think of bolting the loads to the batteries as a ridiculous dinosaur process.

We don't bolt loads directly to the panels, and we shouldn't bolt them to the batteries either. Just my $0.02.

[Raul]

Some well designed systems are actually wasting very little power, only very few. And some of that equipment already exists; it takes a little know how and imagination to put things together even if they were not designed in that way but your creativity can give them a new integration. Think of a voltage sensing relay with a sufficient delay time can turn a load on at the end of absorb. I have a hub that communicates with all components: charge controllers, inverter, dc loads, shunt, batt monitor, AC consumption and I can input various settings to utilise most of the electricity. I can even control it remote anywhere on earth with internet connection.
But I agree, would be nice if the charge controller could communicate with a shunt battery monitor and continue the harvest once the battery is full, to direct it elsewhere.

[AzRoute66]

[...] But I agree, would be nice if the charge controller could communicate with a shunt battery monitor and continue the harvest once the battery is full, to direct it elsewhere.

Yes, that would be an indirect way of accomplishing much the same thing. I like my solution better as it allows for not only monitoring, but actual control of the battery.

[Mike90250]

With off grid systems, there is a trade-off between over paneling to increase charging in poor conditions, or having to run the generator. The optimum ratio varies by site and user lifestyle.

With solar, when the batteries are full, the charger simply throttles back to keep a basic float charge, no extra heat.

[PNjunction]

Um, it's not the controller wasting power, it is just the batteries terminal voltage is slowly reaching the controller's CV voltage limit, and therefore with less and less voltage differential between the battery and controller, less current flows. Even the simplistic PWM types operate this way.

Although one can say that pwm is an on / off switch, it is doing so at 300khz or so. Don't confuse PWM in general with the 70's style of "ping pong" controllers, that have an upper and lower voltage limit, and use battery hysteresis as the inefficient charge mechanism.

But yes, many have wondered how to put this wasted power that the *panels* are capable of when the system has reached float etc. Thing is, the sun sets on you before the batteries are *truly* fully charged anyway - at least with agm.

Thoughts: If you reach float early in the day on your main battery system, perhaps run with a secondary controller from the panel serving an auxiliary non-critical battery that you'd like to keep charged as well. Something that doesn't need cycling so you can dedicate most of your power to your primary battery daily, but just pump in some excess float to an auxiliary battery ...