Skip to main content ↑ ↓
Topic: Interfacing wind turbines to the grid (Read 835 times) previous topic - next topic

Interfacing wind turbines to the grid

Custom-building wind turbines... is hard.  Interfacing wind turbines to the grid... is also hard.
But it is lucrative to come up with a way to grid-tie an inverter, sell the excess energy when the batteries are full.

There are products you can buy, but when they fail things can go drastically wrong.  They're all made in China, of course.  While I'm seeing better products from China all the time, the flood of bad ones hasn't disappeared.  This example might be a good one: https://www.engelecenergy.com/On-grid-Inverters-pl3261504.html
But it doesn't charge batteries.  I want to do both: Keep a battery set charged, and export excess energy to the grid.

My solar panel hybrid inverter does that.  Why can't my wind turbine do that?  I've checked, and my SOL-ARK inverter will not manage the operation of a wind turbine.  It will not accept 3-phase inputs nor will it accept inputs that drastically vary from 50-60 Hz.  While I've been looking for a way to interface my wind turbine to the rest of my RE system, so far all it does is charge the batteries.

A wind turbine bought as a commercially-available unit will usually includes a grid-tie interface, unless the WT is designed solely for charging batteries or running pumps, etc.  Examples include Bergey and Proven, where you can choose from either on-grid or off-grid versions.  Again, if you want both, you need to get more equipment.

Also, their GT inverters are designed for the specific wind turbines they build and sell, not "any" WT.  Since I'm custom-building a wind turbine, one of these would require substantial modifications to adapt it.

What am I to do?

Improvise, I suppose...
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #1 –
First I have to figure out exactly what I want to do, then out how to do it.
First: What to do:
1) Export energy to the grid?
2) use the AC for other purposes?
3) use the energy for stored heat?

My PV set-up allows me to do both.  When the PV inverter is exporting, it's not going directly to the grid, yet.  It goes to a common bus at the base of my meter, and if the house is using power then the house can be supplied by PV.  This is passive and just the balance of current flows, based on Ohm's law, and it's marvelously simple.  I don't have to distinguish between goals 1 and 2, then.  Goal 3 is much harder, since I put the PV panels on the garage, not the house, meaning storing heat in water or sand doesn't let me heat anything that needs to be heated.

Actually, the excess energy from my wind turbine is being used for heat - just a simple bank of resistors that lose heat to the air in my garage.  Hardly useful but necessary to ensure there's always an electrical load on my wind turbine.
How can the excess energy be better used?
1) feed wind DC to the batteries that back-up the PV inverter?
2) feed the WT's output into a separate grid-tie inverter?
3)

Number (1) is obvious, since my wind turbine is designed charge batteries, and that's exactly how I have it hooked up.  But can the system export to grid the extra wind energy?  No.  The problem with (1) is that inverters simply DON'T export the charge from batteries to the grid - they are designed to do the opposite: store energy from the grid in the batteries. Only when the grid fails do they withdraw energy from the batteries, and then it's only to run the backup circuits.  This mode of operation is familiar to us in uninterruptible power supplies and off-grid power setups.  So as the wind charges the batteries, they get full and the excess wind energy just gets lost as heat.  At least I have a backup power supply that's always ready.

Solution (2) promises to export excess wind energy to the grid, but the options are limited.  Most grid-tie inverters will simply not accept the feed from a wind turbine.  If you tried, you would find that the inverter disconnects the wind turbine's electrical leads from time to time.  A wind turbine running freely will spin its blades very fast - destructively.

To get anywhere with Solution (2) one needs a separate control for the wind turbine.  It has to be independent of any grid-tie inverter to do its safety-related job.  There are such devices on the market and you can get simple ones or fancy ones.  I already have a controller in my WT system, because it's basically an essential safety device, whether the WT is meant to be on-grid or off-grid.  Almost invariably, a diversion load (resistor bank) goes with it.

With that controller in place, does this help choose the grid-tie inverter?  While I mentioned safety, the device that's connected to the grid really should be a UL/CSA/IEEE tested device.  This rules out most of the products you can find, particularly the blue ones on Amazon.  Aurora used to make a fine machine, but sadly they are no longer in business.

There may be options on the market that I haven't noticed yet, though.  If anyone knows of one, please feel welcome to suggest it!
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #2 –
efficiency losses of around 15-20% but you COULD grid tie the inverter and use a separate charger to charge batteries that is grid powered. My charger is 90% efficient so whatever the grid tie inverter losses would be added to it.

Re: Interfacing wind turbines to the grid

Reply #3 –
I'm feeling like my brain has gone to rot without fieldlines to challenge it.  It's good to read a post that makes me think. 


Here is my feeble attempt at a solution.  It is in no way correct or advisable, but might actually work. 

Have your wind turbine charge batteries, like it does right now.  Use a dump load battery charger to spill the excess power, but instead of a resistor bank like we both have, use a grid tie inverter (simple, maybe micro inverter) hooked to the dump load instead.  There would be some confusion of the mppt tracking of the inverter, but otherwise i think it should work. 

There was a guy (toysforwatts) i think on youtube.  He did a bunch if chinese turbine testing.  He used a small battery and a grid tie inverter (very cheap one)  to export the energy back.  I think it was just battery voltage that triggered the export, and the battery was only big enough to smooth out the surges in power.  It boiled a lot. 

Re: Interfacing wind turbines to the grid

Reply #4 –
Quote from: bigrockcandymountain – on
Have your wind turbine charge batteries, like it does right now.  Use a dump load battery charger to spill the excess power, but instead of a resistor bank like we both have, use a grid tie inverter (simple, maybe micro inverter) hooked to the dump load instead.  There would be some confusion of the mppt tracking of the inverter, but otherwise i think it should work. 

LOL
Consider your cobwebs shaken out!  I left just enough tantalizing hints that you picked up on my train of thought, didn't I?  A very similar thing has been on my mind lately.
I have progressed from a basic circuit analysis, to a diagram, and now I'm starting to consider the code I'd need, because I'm going to use a microcontroller as the brains of it.

The Sol-Ark has many inputs: PV, Battery, Generator, and Load.
Actually, Load is bi-directional, meaning it can supply 240VAC back-up loads and it can accept inputs too.  You can even connect your generator to the Load terminals, even though it's got a Gen hook-up right beside it.  Load functions like like it's another grid-tie interface.  It's designed to allow micro-inverters to be hooked up, expanding the possible PV array options connected through the Sol-Ark to include combinations of either or both DC and AC solar panels.

So that got me thinking...
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #5 –
So what is the best plan you have so far? What exactly would the microcontroller do? You said the load terminals on the solark can take ac or dc is that correct? 

I'm excited for you, and also interested in using an ev as a dumpload here, so some stuff might be applicable to that. 

Re: Interfacing wind turbines to the grid

Reply #6 –
I spent more time this weekend working on the design.  I've attached a circuit below.

A few things to know before going in (not really meant for you, BRCM, but others reading this later):
PV panels have a typical Current-V-Voltage curve.  While each PV panel has a slightly different curve, in general they all look the same: flat top, curve, steep slope down.  Where it curves down, I refer to it as the "knee" if the curve.  You've probably heard the term before, but if not, think about how your knee looks... take a look at the example I found.
MicroInverters are designed to convert PV panel DC electricity into AC for the grid, and connect automatically.  This is super-convenient and simple to install, but let's note that they won't convert DC from just any source.  Most if not all of them also are programmed with a Maximum Power-Point Tracking algorithm.  They will adjust the current flowing at any given moment to find the peak power.  If a microinverter was connected directly to a battery, it would not detect a knee of a curve, because batteries do not have a power curve like a solar panel does, and a microinverter would then either run the battery flat or set itself on fire.  Place bets on which fails first.
Wind turbines are built either for connecting to the grid or connecting to batteries.  Generally the battery-connected wind turbines are for off-grid purposes, and require a battery to work.
Batteries are a great part of any RE system, but very rarely can they used as an energy storage for exporting energy to the grid.  Some inverters can do this, but I have found that the conditions needed to allow an advanced hybrid inverter to export energy stored in a battery would make it impossible for the battery to be charged by anything other than that advanced hybrid inverter. Usually people who want their batteries to backup their AC will either use an off-grid inverter or a hybrid inverter but allow no other connections to charge the batteries otherwise.

At this point, it seems like a battery-charging wind turbine is therefore not going to be grid connected.  End of story?

...but what if...

I'm working on a device that can unite the battery to the grid, thus exporting excess energy that would otherwise be spent as heat in the diversion resistors.  It's done via a microinverter, but also controlled to trick the microinverter into thinking it's connected to DC solar panels.

In the circuit diagram, the charge controller, diversion load resistors, micro-inverter, and relay are all common off-the-shelf parts with no modifications.

The Arduino/Feather has several functions.  If you've ever seen an Arduino or a similar micro-board, you have seen you can stack boards together on the same microcontroller board.  For this I'll be putting a "relay board" on top of a "prototyping board" on top of a Feather/Arduino board in a 3-layer stack.  I'm not sure but I may need to do something strange to get the FET's to fit on the prototyping board before stacking the relay board on it...  I haven't built these parts yet so I don't have pictures yet.
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #7 –
At this point, the ideas I'm sharing are partly based on a device I developed with Glen Hurd (aka GHURD) from the old Fieldlines board.  He deserves credit for doing a lot of development work to make a similar device work, which I built and used for years.  It was called an "opportunity diversion load timer", which "sniffed" when the charge controller was running the diversion load, and switched on some kind of DC or AC load that could be used instead.  While the AC or DC load was running, the batteries were being drained thus the need for a diversion load by the charge controller was relieved.  So instead of running a heater in my garage, a heater kept water for our horses thawed in the winter.  This saved me a lot of money and this device paid for itself quickly.

I just wanted to pause and give a shout out to GHURD, wherever you are.  Your ideas live on and I'm still using them.
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #8 –
A bit more explanation of how the circuit works:
  • Detects when the Charge Controller activates the diversion load resistors by sensing voltage, then closing the relay contacts. Otherwise, this device is dormant and the relay is Open.
  • Once triggered, the device closes the relay, and stays active for a pre-set time, then shuts off.  If the Charge controller is still running power through the diversion resistors, then the device triggers again for another cycle.  The activity cycle can be adjusted.  Let's assume between 15 minutes and 1 hour, for now.
  • The relay only supplies + battery voltage to the DC positive terminal of the microinverter. 
  • The DC negative terminal of the microinverter is grounded through a pair of power resistors, totaling 4.5 Ohms.
  • Even so, the microinverter should begin to activate as soon as the relay closes.  At first the microinverters will not pass any DC current, just measure DC voltage, so it won't see a voltage drop across the power resistors, yet.
  • Microinverters always have a 1- or 5-minute delay before they actually connect to the AC, so the device will just wait in the meantime.
  • While waiting, the gate driver is already active.  What it's doing is cycling the FET on and off across the 4-Ohm power resistor.  I'll explain the reason for this later.
  • Once the Microinverter comes on-line, it will simultaneously connect to DC, admit current through the DC lines, connect the AC output lines, and  start seeking the MPPT point.  Since it's pulling DC through some power resistors, it will detect a voltage drop.
    This is intended because solar panels have a strong voltage drop while current increases.
  • Unfortunately, if only power resistors were used, the MPPT algorithm in the microinverter would measure higher power at ever-increasing current.   Is 10 Amps enough? 10 Amps X 40V is 400W, great.  Is 11 Amps enough? 11 Amps X 39V is 430W, better... and it keeps doing this, until it exceeds its own power rating (and/or catches fire).  Microinverters are designed for solar panels only, and the "knee" of the curve keeps them safe.  I don't believe that the microinverter will self-limit (but maybe it can).
  • To keep the microinverter from blowing itself up, I've introduced a FET shorting across the 4 Ohm resistor.  By doing this, the voltage drop can be controlled and simulated, like a PV panel was there.
What does the FET do?
  • When the FET is on, the current is basically shorted to ground, which is like the typical use of most N-channel MOSFET controls for motors and heaters.  The microinverter will see lower resistance and current can increase.  Left to its own, its MPPT will seek more current to increase power output.  This simulates being on the right side of the knee of the PV Panel I-V curve, increasing current and voltage will drop slightly.
  • When the FET is off, the current admitted by the microinverter is subject to a lot of voltage drop, so much that there is a strong voltage drop.   Now the MPPT algorithm will measure lower voltage and determine that the input power is dropping, thus the current is too high.  This simulates being to the left the knee of the PV Panel I-V curve, where voltage has dropped due to more current.

This the Arduino needs to control the gate of the FET seeking a balance between current and voltage.  This will need tuning so that the characteristics of the microinverter don't set up a "fight" between the microinverter's MPPT algorithm and the Aurduino program I write to manage the same thing.
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #9 –
Quote from: SparWeb – on
At this point, the ideas I'm sharing are partly based on a device I developed with Glen Hurd (aka GHURD) from the old Fieldlines board.  He deserves credit for doing a lot of development work to make a similar device work, which I built and used for years.  It was called an "opportunity diversion load timer", which "sniffed" when the charge controller was running the diversion load, and switched on some kind of DC or AC load that could be used instead.  While the AC or DC load was running, the batteries were being drained thus the need for a diversion load by the charge controller was relieved.  So instead of running a heater in my garage, a heater kept water for our horses thawed in the winter.  This saved me a lot of money and this device paid for itself quickly.

I just wanted to pause and give a shout out to GHURD, wherever you are.  Your ideas live on and I'm still using them.
SparWeb;
He's still kicking. I have a few conversations with him and Tom W, from time to time.
I still have the ones I purchased from him too. I'm currently redoing my little 200Watt solar setup to try out a smallish sand heater setup for our basement media/playroom.
 This looks to be a nice setup, like BRCM I'm curious where you go with these.

Cheers
Bruce S

Re: Interfacing wind turbines to the grid

Reply #10 –
Quote from: Bruce S – on
This looks to be a nice setup, like BRCM I'm curious where you go with these.

Well, slowly, is where I'm going.  Diving into the circuits, I'm afraid of how choppy the signals are to monitor current.  I'm adding various things to smooth it out, so that I don't need a complicated program in the software.
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine

Re: Interfacing wind turbines to the grid

Reply #11 –
Quote from: SparWeb – on
Quote from: Bruce S – on
This looks to be a nice setup, like BRCM I'm curious where you go with these.

Well, slowly, is where I'm going.  Diving into the circuits, I'm afraid of how choppy the signals are to monitor current.  I'm adding various things to smooth it out, so that I don't need a complicated program in the software.

Feed it thru a rectifier and monitor the DC current via your micro... a full wave bridge with electrolytic has 1.4 times the voltage output so some math involved...

Re: Interfacing wind turbines to the grid

Reply #12 –
I looked at your circuit diagram and read the explaination, and I'm trying to picture exactly how it would all work and i think i see some light.  I have "Electronics for Dummies" open in my lap right now.  That helps too. 

I have something I'm wondering about and I'm sure someone will explain to me why it won't work. 


In my case, my diversion load is a 500hz 12v signal that goes to a ssr and dumps the 48v battery voltage through an approx 10ohm resistor (water heater element) I assume the 48v comes out of the ssr at 500hz, not exactly as dc.

If I swapped the water heater element for a grid tied micro inverter it would get fed messy 500hz and would blow something up likely. 

What if all you did was add a very large capacitor (or a few of them) with a resistor in series to limit charge /discharge amps between positive and negative between the ssr and the micro inverter? Would it smooth the dc power enough that the micro would be happy? Would it store enough that when the mppt swept up it would drop a bit of amps and mimic a solar panel?

I get the feeling the magic smoke would come out somewhere, and i don't even have the grid here, so it's all academic.  I just can't picture in my head what would happen, and i was hoping to learn something by asking.

It sounds like a thread hijack when I read through it, but it kind of applies to sparweb's setup too. 

Re: Interfacing wind turbines to the grid

Reply #13 –
Quote from: MaryB – on
Feed it thru a rectifier and monitor the DC current via your micro... a full wave bridge with electrolytic has 1.4 times the voltage output so some math involved...

For now, I'm monitoring DC current, which is pulsed.  Where I tap off the power resistors with a voltage divider determines the scaling factor I need to satisfy the microcontroller's ADC input.  I figured out that a 1 to 10 uF capacitor is all I need, and doesn't have too slow a time constant for these devices.
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine
 

Re: Interfacing wind turbines to the grid

Reply #14 –
Check this out:
https://www.falstad.com/circuit/

Draw your circuits, watch them work in real-time.  Or slowwww it down at count the 1kHz pulses one millisecond at a time.
6.5 kW PV solar array + 12k Sol-Ark Inverter
10ft (3m) diameter custom-built wind turbine