Optimizing MPPT Curve for Wind Turbines

[Jan Ben]

Hi,
Just wondering if anyone has tried it before or what are your opinions for directly connecting a variable resistor bank to the output terminals of an MPPT controller?

In this scenario it will be something like:
1. Wind Turbine Output from Rectifier to the +/- INPUT terminal of charge controller.
2. 12V battery to the +/- SENSE terminal of the charge controller.
3. Variable Resistor Bank to the +/- OUTPUT terminal of charge controller?

MPPT have data logging functions for input and output power, voltage, and current coming into and out of the charge controller.
So I was wondering if I can set an initial MPPT curve and then vary the resistance to control the voltage at the input and output power (to some degree up to the max in the initial curve).
And so I get the power output values at somewhat near constant voltages and optimize the MPPT curve from there?

Any thoughts?

Thanks.

[Rob Beckers]

Probably not a great idea Ben.

MPPT charge controllers needs to see battery Voltage to function. In fact, when you look at installation manuals they all say to connect the battery side first, then the PV/Wind side (and the opposite order when disconnecting). The risk is that the charge controller gets blown.

In theory you could set an MPPT curve that under-loads the turbine, and add a variable resistor to add additional load to the input side of the charge controller (so between positive and negative). Turbine output would be the sum of the charge controller plus what's going into the resistor. That could/would find a maximum, though it's probably not all that useful.

The issue is that you have to relate the maximum to Voltage, and changing the load also changes the Voltage that the turbine runs at.

The "official" method is to have an anemometer somewhere near hub-height, and to log power output vs. wind speed. Start with an MPPT curve, run for a while, and plot power vs. wind speed. Then increase the MPPT curve by (say) 20% and repeat. Decrease MPPT curve by 20% and repeat. That gives you a set of power curves showing wind speed vs. output power, and the goal of any MPPT curve is to optimize that (most output power for each wind speed). From the trends in those curves (which one is better) it is pretty clear in which direction the MPPT curve needs to move, and you can tell this for each bracket of wind speeds. So, rinse & repeat until the best MPPT curve has been found.

-RoB-

[Jan Ben]

Hi Rob,

Thank you for the reply.
Yes, I'm currently using anemometers to record wind data both near hub and free/surrounding wind.
I did some initial test and ran into some problems.
The MPPT charge controller seems to be limiting the output power going into the battery (around 60-80W) no matter how much I increase the MPPT curve settings.

These are the test MPPT curve settings for the attached images.
In the images, the red shows power analyzer readings of input power going into charge controller sampling at every 0.5s (binned/averaged to 1 sec), the blue shows the MPPT charge controller output readings at 1 sec and yellow is power analyzer reading of generator voltage.
MPPT Curve 1:
Voltage - Watts
0 0
12 20
14 40
16 60
18 80
20 100
22 120
24 140
26 160

MPPT Curve 2:
Voltage - Watts
0 0
12 400
14 500
16 700
18 900
20 900
22 900
24 900
26 900

It's my first time working with charge controllers and batteries in general.
But based from the graphs I made while analyzing results, clearly the generator is spinning at greater than 12 volts but is only outputting 100w or less to the battery.
Does battery capacity (Ah) affect the power required to charge it (output power of charge controller)?
I didn't have any diversion load on my setup and I am still using only 40Ah battery for this test.
How does Ah relate to the Charging Watts?

Thank you.

[Rob Beckers]

Have look at the battery Voltage Ben. Maybe it's pushing up to bulk/absorb Voltage and the charge controller limits output for that reason?

-RoB-

[Jan Ben]

Hi Rob,

Thanks for the reply.

I see, that actually may be the case because I was not using any diversion loads in parallel with the charge controller output.
One thing is that, battery capacity was very small as well 40Ah, this is also a factor right?

I am confused whether it would be better (less complicated) and easier for optimizing the mppt curve to just increase battery capacity or add the diversion loads (after charge controller output).

If I add dump loads, I might size it too much or too little and it would seem to be a very manual and discrete process of trial and error on # of dump loads to add.

How will I know that if I increase the battery bank capacity to X Ah, it's can go up to 300 or 400W when charging at 13-15V?

Does anyone know or can point me to a resource on how to determine how fast the battery voltage increase when charging?

[Jan Ben]

Sorry I can't seem to edit my last reply.

Just another thought: would this work - using an electronic load operating in constant voltage mode instead of gradually adjust mppt curves of charge controllers?

[Rob Beckers]

That's what I was going to suggest, you beat me to it: A load that keeps the Voltage of the battery below the bulk/absorb Voltage setting. During bulk charging the charge controller will not limit the current coming from the turbine, it'll pump as much as it has into the battery.

If you're using a lead-acid battery there may be the downside of killing that battery over time. Lead-acid batteries need to see a full 100% on a regular basis (every 2 weeks or so) to avoid permanent sulphating. Though if you start out with a fully charged battery, you could keep the Voltage around float (13.6V for lead-acid). That will keep the battery charged, while hopefully still tricking the charge controller into bulk charging.

-RoB-