5

I am trying to understand how MPPT charge controllers work when the load is less than the solar power generation. Let's assume an off-grid system with battery backup and no dump load. In my understanding, the following situations can happen:

  • solar < load: solar will partially fulfill load requirements, rest will come from the battery
  • solar = load: all of the load is fulfilled by the solar directly, battery status remains unchanged
  • solar > load (battery not fully charged): solar will power the load and rest will go to the battery.
  • solar > load (battery fully charged): In this case, the charge controllers curtail the power generation to match it exactly with the load. The excess solar energy is lost in the form of heat.

I want to understand how this curtailment happens in actual MPPT charge controllers. Do the MPPT techniques/chips leave the MPPT mode and the operating point of the solar panel is controlled by the current drawn? In that scenario, how does this whole circuit (DC-DC converter) knows how much demand is out there? Is there any implicit feedback loop?

I will highly appreciate your detailed response on the question. Please let me know if the question is ambiguous.

Noman Bashir
  • 51
  • 1
  • 3

4 Answers4

2

MPPT controllers are not about power balancing, but power maximization.

Let's say you have a system like this:

  • Solar panels ⇒ MPPT controller ⇒ battery charger ⇒ battery ⇒ load

If the load on the system is smaller than the incoming solar power, the battery voltage will rise.

When the battery voltage reaches a point where the battery charger decides that the battery is full, the charger will cut out, and stop drawing power from the MPPT controller. In other words, the battery charger will present a high impedance to the MPPT controller.

The MPPT could react to this by presenting a high impedance to the solar panels, effectively stopping any significant power draw from the solar panels.

When the battery is not being charged, the load will cause the battery voltage to fall. Then the charger will start drawing power from the MPPT again, and the MPPT will start behaving like an MPPT again. And the cycle repeats. The system can cycle back and forth between these two states several times a minute, or even thousands of times a second, all depending on the time constants (battery size or capacitance) and hysteresis in the system.

You may have a commercial product called an "MPPT" that has more than an MPPT controller. For example it could have an integrated battery charger and maybe even an integrated voltage regulator for the battery output. Don't let any oversimplified marketing jargon confuse you. The MPPT part of the device only handles MPPT, and if the device also actually charges a battery in a proper manner, then it must also have a battery charger built into it.

Dampmaskin
  • 3,787
  • 1
  • 16
  • 29
  • Thanks for the response, it clarified a lot of things. However, I have another question for you. Let's assume the system looks like this: Solar panels ⇒ MPPT controller ⇒ inverter ⇒ load Assuming the solar > load, how would this system behave? – Noman Bashir Aug 30 '18 at 21:34
  • 1
    Then feel free to create a new question. – Dampmaskin Aug 30 '18 at 21:35
  • I am sorry, I pressed Enter by mistake. I have added my new question in the previous comment. It's related to the first question, so I want to avoid creating a new thread. – Noman Bashir Aug 30 '18 at 21:36
  • I believe I answered this in the two paragraphs starting with "When the battery voltage ..." But I will add a third paragraph to describe how a system might behave over time. – Dampmaskin Aug 30 '18 at 21:37
  • I do not have a battery in the new scenario. Assume, I am trying to run my load directly of solar panels and my system looks like this: Solar panel --> MPPT Charge Controller --> Inverter --> Load I know it's not commonly used and maybe not advisable. However, I am interested in understanding how would this system work theoretically. Note: I do not have a battery in this system. – Noman Bashir Aug 30 '18 at 21:41
  • Your original question did describe a system with a battery. Just saying. But even without a battery you could probably use a pretty similar model, just pretend that the battery is extremely tiny. Or maybe you need to add a big fat capacitor to the MPPT output for stability. It depends. – Dampmaskin Aug 30 '18 at 21:46
  • As most of the inverters allow a range of input DC voltage, would we still need a capacitor in that case? As far as the working is concerned, please let me know if my understanding explained below is correct. In this scenario, less load will mean high impedance at inverter's output. The inverter will, in turn, present a high impedance at its input (output of MPPT charge controller), which will propagate it further to the solar panel and less power will be produced (as solar panel's operating point on IV curve will change based on new impedance). However, how MPPT comes algorithm works then? – Noman Bashir Aug 30 '18 at 21:55
2

Your assumptions are correct, except:

solar > load (battery fully charged): In this case, the charge controllers curtail the power generation to match it exactly with the load. The excess solar energy is lost in the form of heat.

It's not converted into a heat, rather not all available power from panel is being converted.

THe MMPT seaks the optimum point by measuring voltage and current from panel and then multiplying to get the power. With rising/decreasing the output voltage of the DC/DC converter the current trough panel is changing, thus the MPPT point can be found.

If the "demand" is less than available solar power, then the output voltage of the DC/DC converter will rise to its maximum allowable voltage - probably set as a parameter on a controller. Even if the voltage is high and no suitable load is present, then there will be just a minimal current drawn from panel, of course out of MPPT point, but this can't be avoided. The most important thing is that the PV voltage does not fall too low when the demand is graeter than available supply. In such way you get the MPPT point.

Marko Buršič
  • 23,562
  • 2
  • 20
  • 33
  • 3
    _”It's not converted into a heat”_ It’s a stretch. The PV panel will become hotter. – winny Aug 30 '18 at 21:42
  • @winny For example let's say that the panel is completetly unloaded, current iz zero - why it would become hotter than it would be optimaly loaded? – Marko Buršič Aug 30 '18 at 21:48
  • 1
    https://electronics.stackexchange.com/questions/313804/excess-energy-from-my-off-grid-system this post suggests that the energy from sun that is not supplied to the load is lost in the form of heat. – Noman Bashir Aug 30 '18 at 21:49
  • @NomanBashir Interesting, but I haven't found an exact statement about that in the linked reference. Have you? – Marko Buršič Aug 30 '18 at 21:59
  • 4
    The energy has to go somewhere. If it's not reflected as EM radiation, I can't see where else it could go, except heat. Are open circuit PVs more reflective? – Dampmaskin Aug 30 '18 at 22:07
  • Basic physics. All energy input must go somewhere and if it isn’t lead away as electricity and used elsewhere, all the nature have left is to increase the temperature until it’s radiated away. – winny Aug 30 '18 at 22:07
  • 1
    @Dampmaskin In the medium to deep IR spectra you should be able to see the difference with a camera. – winny Aug 30 '18 at 22:08
  • OK, I buy it. It makes sense, I haven't seen panels changing coulour, but I didn't seen any freezing, also. – Marko Buršič Aug 30 '18 at 22:10
  • @winny That is interesting. If the change is immediate as the circuit is opened and closed, that would imply reflection. If it's slow, that would imply thermal radiation (i.e. heat). Agreed? – Dampmaskin Aug 30 '18 at 23:10
  • @Dampmaskin Sure thing! – winny Aug 31 '18 at 19:39
1

Do the MPPT techniques/chips leave the MPPT mode and the operating point of the solar panel is controlled by the current drawn?

Yes. The panel voltage will rise when the battery is fully charged.

In that scenario, how does this whole circuit (DC-DC converter) knows how much demand is out there? Is there any implicit feedback loop?

The MPPT controller contains two feedback loops: One regulates the input voltage, keeping at the max power point, and the other regulates the output voltage. If the output voltage is below the setpoint, i.e. the battery is absorbing all available power, then the first loop is used. When the output voltage reaches the setpoint the second loop controls. Otherwise the battery would be overcharged.

τεκ
  • 4,229
  • 1
  • 16
  • 17
  • Hi, thanks for the response, it makes a lot of sense. Would you mind answering this question? Assume, I am trying to run my load directly of solar panels and my system looks like this: Solar panel --> MPPT Charge Controller --> Inverter --> Load. I know it's not commonly used and maybe not advisable. However, I am interested in understanding how would this system work theoretically. What would happen if the load is less than what the solar panel can generate? – Noman Bashir Sep 01 '18 at 01:34
  • @NomanBashir If the load is less than the maximum power of the solar panel, the load voltage will rise to the setpoint, then to prevent the load voltage from going higher less current will be drawn from the solar panel and the panel voltage will increase above the maximum power point. The MPPT charge controllers I've seen specifically warn against using without a battery saying it will damage the controller. – τεκ Sep 01 '18 at 02:00
0

Do the MPPT techniques/chips leave the MPPT mode and the operating point of the solar panel is controlled by the current drawn?

Yes, of course. Low load means less power is taken from the solar panel. If you look at the IV curve of the solar cell below (courtesy of wikipedia) you'd see that with reduced current the cell will rise almost to it's maximum voltage and stay there, way below MPP.

enter image description here

In that scenario, how does this whole circuit (DC-DC converter) knows how much demand is out there? Is there any implicit feedback loop?

Yes, any DC-DC controller with configurable output voltage has a feedback circuit. And no, it does not "know" the demand. It does not have to. The purpose of feedback is to keep output voltage at given level.

The DC-DC converters are basically energy transfer devices. They regulate the power (which is the rate of energy transfer) by changing frequency and/or duty cycle. Note that consequently this also defines input current (and answers your question "how this curtailment happens"). Let's compare different devices for clarity.

  • Simple DC-DC controller uses feedback of the output voltage to adjust power so that output voltage stays at required level regardless of current drawn by the load. It does not care about power source, drawing as much power as necessary or available.

  • MPPT controller uses input voltage and current sensors and adjusts power draw so that solar panel stays at MPP. This works best with loads that have wide operating voltage range or capable of adjusting to whatever power is available to them. Some examples are water heaters, water pumps and inverters. External battery chargers can be used as well. With careful selection of panel size even primitive over-current circuits could be sufficient for some battery chemistries.

  • Finally, most complex device is MPPT charger controller. It uses sensors on both input and output to ensure correct CC/CV charging profile while at the same time keeping panel at MPP. This, of course, works mostly during power-hungry CC charging stage. During CV stage output current drops significantly enough to force input out of MPP mode.

One very important thing to note, that the above does not account for load connected in parallel to the battery. I could not find any MPPT charge controller with built-in power management (doesn't mean there aren't any, of course). What this means is that using charger controller with load is not very good idea, since it cannot distinguish current consumed by the load from current going into the battery.

Better option would be to use two-stage circuit, with MPPT controller taking care of extracting maximum power on the first stage and charger with power path capability taking care of the most efficient distribution and correct charging profile on the second stage.

I strongly recommend reading SLPA013A by TI. It has algorithms for exactly this kind of setup, although implemented in rather complex system.

Maple
  • 11,755
  • 2
  • 20
  • 56
  • What if we are trying to serve only a load, no batteries, with an MPPT charge controller? Let's assume, my system looks like this: Solar panel --> MPPT Charge Controller --> Inverter --> Load. I know it's not commonly used and maybe not advisable. However, I am interested in understanding how would this system work theoretically. What would happen if I want this system with a conventional MPPT charge controller? Or what changes would be needed in an MPPT charge controller for this system to work? – Noman Bashir Sep 01 '18 at 01:40
  • Why would you use _charge_ controller with anything but the batteries? Of course it won't work normally, because controller will try to regulate load current using battery charge profile. Most likely one of the safety features will kick in and shut it down, thinking there is something wrong with the battery. I don't think you can do anything to make this work. It's like trying to grind coffee with a blender. You might be able to do it eventually, but why bother? Especially considering that MPPT controllers are cheaper than MPPT charge controllers. – Maple Sep 01 '18 at 04:42
  • Let's assume that I have built my own DC-DC controller with MPPT algorithm. This controller does not assume a battery at the load and simple attempts to maintain a fixed voltage at the output of the DC-DC converter. I don't want to use a battery and power any load (below maximum solar power). What would the challenges be? – Noman Bashir Sep 04 '18 at 19:22
  • As I explained in the answer, the MPPT controller works best when load can adapt to available output power. So, I guess the main challenge would be to maintain a fixed voltage at the output. You need to figure out exact PWM parameters for driving the switch, to keep output voltage stable at given load current, while also keeping panel at MPP. Not to mention that MPP itself constantly changes with light conditions. However, you've mentioned _inverter_ above. Inverter does _not_ require fixed voltage, which is exactly what makes it suitable as a load for non-charging MPPT controller. – Maple Sep 04 '18 at 19:54
  • Thanks, this cleared a lot of things. However, I don't understand why do we need to keep the panel at MPP? Assume, I have this complex MPPT charge controller, that keeps the output voltage constant while keeping the panel at MPP. However, if this (load_current x fixed_voltage) < MPP, the panel should operate at some point below MPP, correct? – Noman Bashir Sep 04 '18 at 20:17
  • See the first paragraph in the answer and the graph following it. – Maple Sep 04 '18 at 22:33
  • Great! This point seems not very well known to most. I am also intensely interested in either a "pure" MPPT converter, or a DC-load-serving MPPT grid-tied inverter. Do you know of any? Basically this just requires converters with the DC-bus architecture. – Milind R Mar 08 '19 at 11:37
  • @MilindR Sorry, I don't know much about commercial devices on the market. I was trying to explain the theory of MPPT operation here, hoping it would help with practical implementations. – Maple Mar 09 '19 at 16:25