Power curve problems

[Charles Servaes]

Hi I'm new to your site. I am having problems with the power curve settings for my HYE-3000 watt wind turbine.I have a ginlong GCI 5 k wind and solar grid tie inverter. The wind turbine is 45 ft off the ground. I put a power curve setting into the inverter which came from HYE as 70 volts=120 and the top end was 320 volts =3090,When I did that the wind turbine would only produce around 90 watts in 12-15 mph winds. The wind turbine moves very slow as if the curve setting was to high and the turbine will not produce 120 watts at that slow of a speed.Hope someone can point me in the right direction to get the most out of my wind turbine.Thanks!

[Rob Beckers]

Hi Charles,

How do you measure the wind speed at hub-height? Do you have an anemometer up on the tower?

To calculate an MPPT curve I need a couple of parameters (and their accuracy determines how accurate the curve will be). First one is the exact diameter of the turbine rotor (so, twice the length from a blade tip exactly to the center of the rotor).

Second one is a point that should be on the curve: I need the (loaded) DC Voltage, the output power it should produce at that point, and the wind speed that this should happen at. Usually rated output is a good place, if the manufacturer publishes that. If you don't have DC Voltage, AC will work too, as long as I know which it is. This should preferably be loaded Voltage, since unloaded can be much higher.

-RoB-

[Charles Servaes]

Hi RoB: Thanks for your reply

Yes I have a anemometer as close as I could to hub height. Here are the specs for the 5 blade Hy-3000 watt wind turbine. Cut in wind speed 5.6 mph. Rotor diameter 3m. Swept area 7.07meters squared. This is the power curve I got from Hy. I don't know if it is loaded ac or loaded dc? The wind turbine is 3 phase wild ac to rectifier to dc then to Ginlong grid tie inverter. I can only presume that it is dc because that is what the inverter is dealing with? I have another question. Will a 5 blade wind turbine produce more power in low wind then a 3 blade? That is one of the reasons I purchased this produce.

7mph wind loaded volts

Volts 70=120 watts Volts 130=510 watts. Volts 190=1200 watts. Volts 250=2160
80=150 watts. Volts 140=600 watts. Volts 200=1320 watts. Volts 260=2310
90=210 watts. Volts 150=720 watts. Volts 210=1500 watts. Volts 270=2520
100=270 watts. Volts 160=840 watts. Volts 220=1620 watts. Volts 280=2700
110=360 watts. Volts 170=960 watts. Volts 230=1800 watts. Volts 290=2910
120=420 watts. Volts 180=1080 watts. Volts 240=1980 watts.

Volts 300=3000 watts(600)RPM
Volts 320= 3090 watts
The inverter will only let me input by intervals of 30 to 210.

[Rob Beckers]

Charles, the 300V DC - 3000 Watt point, at what wind speed is this happening? I assume that's rated output, they should mention a rated wind speed with that.

Voltage would be DC since that's what the inverter works with. You normally don't need more than half a dozen points, there's no reason to put points every 30 Volt. To make a smooth curve it makes sense to put a point about every time output power doubles.

In theory a 3 bladed wind turbine can do just as well (or better in fact) than a 5 bladed one. The more blades a turbine has, the higher the torque (force), but also the lower the RPM. Since power is in essence the multiplication of those two the power value remains the same, it works that way regardless of the wind speed. There's little energy in low wind speeds, and not much to be had no matter how many blades.

In practice 3 bladed turbines tend to work better than 5 bladed ones, though that's mostly a factor at higher wind speeds and RPMs: The more blades there are the more losses due to drag.

-RoB-

[Charles Servaes]

RoB: They post 26.8 ms at 3000 watts. I also see that the wind turbine has trouble with wanting to flury out of the wind even when the wind isn't to 26.8 ms. Tail not heavy enough or tail not tall enough? I have a 20 kw dump load with the turbine to keep it under control if you think I could modify the tail. I live in Kansas where it is sometimes very windy and have also noticed that the wind turbine with the 5 blades has a problem with reaching 320 volts. The invertor has a pre set program already installed into it . I had to go back to it because the one hy gave me didn't seem to work.

[Rob Beckers]

Hi Charles,

The 26.8 is miles/hour (I found the spec sheet), that's 12 m/s.

So.... After running those numbers the conclusion is that they are smoking The Good Stuff: A really good turbine of 3.05m diameter, with pretty good efficiencies for rotor, alternator, and inverter, will produce right around 1,900 Watt at 12 m/s.

They publish a Cp (rotor efficiency) of 0.36, and if I take that as being true it works out to 2,784 Watt coming off the rotor. Still not 3kW. Coming off that 2.7kW is the alternator efficiency, inverter efficiency and idle losses. On top of that, this turbine runs at a TSR (Tip-Speed-Ratio) of just 4 (likely because of the many blades) and that will hurt efficiency even more. In short, 3kW at 12 m/s is a pipe dream.

So, to go back to reality, this would be a table of wind speeds vs. output using reasonable efficiencies:


Wind (m/s) RMP DC Volt Watt Out
2.5 63 70 0
3 75 84 0
3.5 88 98 6
4 100 111 29
4.5 113 124 57
5 125 137 92
6 150 162 187
7 175 187 317
8 200 210 490
9 225 234 710
10 250 256 983
11 276 278 1315
12 301 299 1712
13 326 319 2178
14 351 339 2720
15 376 358 3343
16 401 376 4053
17 426 393 4855


An MPPT table from this would look like this:

110 Volt - 0 Watt
125V - 60W
160V - 185W
210V - 490W
260V - 990W
300V - 1700W
320V - 2200W
325V - 3500W

That last point is to just put the brakes on the turbine, get as much out of it in high winds as we can (before it furls) and because the manufacturer says the alternator is good for 3.5kW.

This should be a good starting point for a working MPPT curve. You can optimize it since you have an anemometer: By logging wind speed vs. output power. The idea is to move the MPPT curve values by (say) 20%, make another wind speed vs. power curve and compare, you want that curve to be as high as possible (most power for each wind speed).

-RoB-

[Charles Servaes]

Dear RoB: Thank you for the time you spent on the power curve. I have put the numbers in and the wind turbine is doing a lot better now,more satisfied with the results. I'm happy that there are forums like this one, that you can talk to people who know about such things. I found out one thing,that most wind turbine sellers are smoking the good stuff.


Thanks again: Charles

[Rob Beckers]

Good to hear it worked out Charles!

[Harri Piltz]

I have been facing similar problems.
Fixed blade, small turbines are really difficult, maybe impossibe to 'optimize' the energy production.
For various reasons:
- good (not even accurate) wind speed measuremen is difficult
- winds tend to be very turbulent (one should be good in predicting the coming speed+direction)
I had a quite interesting article about some kind of MPPT control principle.
See http://cdn.intechopen.com/pdfs-wm/9563.pdf
or
http://yadda.icm.edu.pl/yadda/element/bwmeta1.element.baztech-9fd0d6b1-61c2-4659-b689-a2e4d1b0a5fc/c/Gitano_v14i1_01.pdf

Intelligent electronics makes the 'stuff'.
Very simple, looks good.
Do you think it would be working for any kind of small mills ?

I have been looking for a control system just for heating.
Got a quite good(enough) system from Scoraigwind/Hugh Piggott.
He has a PWM (PulseWidthModulation) control idea for resistors.
He also may have this posting on his website.
http://scoraigwind.co.uk
Do not know.
All the best to you both.
Harri Piltz

[Rob Beckers]

Harri, thank you for the links to the articles! Those are actually quite good. The first one is a great overview of how MPPT for wind should work, and what it does. I've not seen much explanation about this, and had to work it out for myself. When I have a bit more time (maybe over the weekend) I'll read them in detail.

By the way, to optimize an MPPT curve you don't need an accurate anemometer (wind speed measurements), it just needs to be reproducible. What one does is log wind speed vs. output power, and plot an (averaged) curve of this. Then shift the MPPT table up or down 20% and do the same thing again. The new power curve will show if for certain wind speeds the changed MPPT curve works better (or worse). Do that a few times and you'll very quickly get an optimized MPPT curve for the turbine. The wind speed measurements don't need to be absolute, the actual value of the wind speed is not important, as long as the same wind speed results in the same measured value between different measurements. The anemometer would ideally be at hub height, but for most that's a problem (you'd need slip-rings), next best is to put it just below the bottom of the blades on the tower.

At the end of the day it's all about optimizing the turbine's power curve: Wind speed vs. output power at that instant.

If the air is very turbulent (typical for small turbines on too short a tower) you'll get a very wide range of wind speeds that result in about the same output power. In short, a wide spread of measurement points that make it hard to get an (averaged) power curve.

Remarkably very few small turbine makers actually take the time and effort to optimize the MPPT tables that they supply. I've seen some atrocious ones, where it's clear the turbine maker has no clue how this actually works.

-RoB-

[Harri Piltz]

Yes I agree.
The MPPT 'scheme' looked very good to me.
Let the electronics do the 'optimizing work'.

I also understand the importance of relative measurement.

Small mills in turbulent envoronments are quite difficult.

Because both absolute wind speed measurements and relative ones have a very vide uncertainty. Maybe just those +/- 20 %.
Improving/deminishing the uncertainty is (quite) impossible.
So why try to improve the measurement accuracy ?
Which is more or less waste of time.
Real improvement would be to be able to predict the 'gust' coming or going, and adapt the mill output accordingly. This is not a DIY stuff.
In small mills there are many other points to be able to improve the output.
Like TipSpeedRatio, which is fairly well taken care of the electronics in those papers.
There are also other control algorithms than this HillClimbing method, which seemed to me quite simple (or simple enough), and it will save the PWM programming effort, if it would work. I do not know... just have to trust the paper, made sense to me.
All the best.
Harri Piltz

[Harri Piltz]

My initial message, when reading different 'wind- or ecosites', was, that generally people buy stuff without knowing, what they are doining. They trust suppliers, with no or very little knowledge of what their equipment is really doing or how and where to use them.
Specs are often this and that, not to speak about, how those specs fit into just my situation.
Windmills + other 'ecopowerstuff' require quite a big deal of knowledge before they get a full power. There seems to be a lot of believes based on almost nothing concrete.
Unfortunately I do not have any good idea, what to do.
All the best.
Harri Piltz