 Notice. New forum software under development. It's going to miss a few functions and look a bit ugly for a while, but I'm working on it full time now as the old forum was too unstable. Couple days, all good. If you notice any issues, please contact me. |
Forum Index : Electronics : Turbine Performance
  Page 2 of 4   |
|||||
| Author | Message | ||||
| VK4AYQ Guru  Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Pete I for one, am way out of my depth with all this, although very interested, I need to get the lights flashing. The idea of computer control is great I think, but it needs to have a back up redundancy system for when it dies that can take over in a non optimized way to stop the wheels falling off the cart. Like a dead mans brake in a train. Murphy Law of electronics states that things fail at the square of the number of components, and with most electronic doovers being made in China now, that then invokes Gybersons Law. All the best Bob Foolin Around |
||||
dereksoftstuff Newbie  Joined: 14/11/2010 Location: United KingdomPosts: 30 |
Pete There seems to be quite a lot of technical information on the site. I'm impressed. I think Electronics are used in cars and airplanes now, so there's no way of knowing what might be next ... Of course, just measuring wind turbine performance is straight-forward, you only turn your electronics on when you want to log something or view data. That's what my system does now. Whereas an electronic system to improve the performance of a system would need failsafes. If the electronics died then the turbine must be able to continue (or stop) as normal. Just as Bob says. It's plug and play (or unplug and don't play). I'm new to this as well, but I don't believe there is any system out there (DIY or commercial) that will do what is outlined in the previous diagram. Wouldn't you like to be able to measure the performance of your system, and compare it to another system. Wouldn't you like to make a modification to your system, and 'see' the difference in performance. If someone who's using an 'mppt' module with their system, is getting 30% more kwh/pa than you are getting with the same system (like for like etc), wouldn't you be interested. |
||||
GWatPE Senior Member  Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Derek, If you are only talking MPPT then I will challenge you on this. I believe the MPPT tracking in a Skystream 3.7 does just that. You have to part with some serious coin first. My own analogue Maximizer tracks the wind energy within a few % across a windspeed range of 2-10m/s. I also had a micro version on a second windmill with the same size rotor in close proximity, and it produced as close as measurable, the same power, and the rpm tracked the windspeed. I was able to confirm that both systems were maximizing the power extracted from the available wind. If you also include logging, then the skystream has that as well with a PC XBee wireless comms. I have one of the micro's you are using on order. When I have had a bit of time with it I will see if my own methods can be translated to it. I still believe that the pic/picaxe chips could be a logging interface as well as a MPPT function, but I believe that keeping these aspects separate is better. Before anyone has a go with a programmable controller, there must be adequate windmill safety protection systems in place. For a Buck converter, if the switching device goes short, then the mill will probably just stall out, but if the device is held OFF then windmill rpm runaway will result and backup measures have to operate. For a boost converter, if the switching device goes short, then the windmill will probably just stall out, and if the device is held OFF then the windmill will still be contained by the main rectifiers. Backup systems may not need to operate. During program changes, the micro is not operating, so safety measures including pullups on output PWM lines, to turn on the switching devices, or some other electrical interlock to stop the windmill needs to be employed. If you have more than an ornamental windmill, and say a 500W and up power rating, then protection systems should be employed. BTW Downwind Pete is right about all the code that I have tried, that did not work. I have also only recently broken and now repaired the power electronics on my buck converter. The maximum power handling requirement of the switching device all of the time with a buck converter will test even proven designs on a windmill. Luckily a mosfet went short, and the battery loading overloaded the alternator and the windmill rotor just stalled. As far as algorithms go, a buck converter controller has to increase the PWM duty and then decrease the PWM signal for increasing windmill power, while a boost controller only has to decrease the PWM duty for increasing windmill power. These are nonlinear functions in both cases. I measure rpm of the rotor at a high rate and with a high precision. Measurement glitches form the main problems. These are most likely caused by noise associated with sliprings etc. It is imperative that the measurement reflects the true rpm. All control loops I tested using current maximizing introduced cyclic loading behaviour and did not give a power output that tracked the windspeed. The loading of a boost converter sees a large rate of change of pulsewidth for a small rate of change of power, at a low available power, so the maximizing PWM system can easily overload the windmill. On my analogue maximizer, I used a voltage controlled current shunt, to vary the current limiting of a PWM chip. This way as the windmill rpm/output voltage rose with increasing windspeed, more current was allowed to be passed to the load. This function was non linear and was able to be calibrated to give a close to cubic power transfer function. Here are some links to some pics I posted on another thread. pic 1 pic 2 pic 3 pic 4 These were pics from Oct 2007. The ccts were refined from Sep 2006, following months of failures. Once I made the voltage controlled shunt, things fell into place and a stable control loop was achieved. The maximizer still operates today, unchanged. The unit has a simple output voltage limiting control at 29.5V, and a small windmill ON/OFF toggle switch brake activator, and indicator for windmill operation and brake ON. The unit is diode blocked from the battery [an inherant feature of a boost cct] and consumes no power from it. The unit does not need to be connected to a battery to still function, and the windmill is still safe, even if the batterey is disconnected. When I made this unit, I was attempting to widen the envelope of wind speeds that my windmill could extract from. I really never had surplus power. I do not try quite as hard now. I have just made my windmills bigger. This has created other problems with what to do with surplus power. I now ensure that as much power is extracted in low winds, and the top end power is restricted by furling and speed limiting. I use grid diversion loading for the bulk of surplus power. I have backup diversion heaters and shutdown mechanisms. The windmills have to operate automatically with solar battery recharging modules. At the end of the day, I agree that we should get the maximum from a windmill. I have altered my perspective to include optimizing for the winds normally seen, and not include the peak times. If we try and maximize for the peaks, then we end up with needing bigger diversion loading systems when surplus power is there. I have limited budget and wall space for equipment, so I now work around the winds I get 80% of the time. This has seen my focus on windmill blades and furling/rotor speed limiting. I don't have a crystal ball but I would want reliability to be high up on my list. Maximizing power from the windmill during typical winds would be there as well. Extending the power range all the way up to the maximum power the windmill could produce is not as important. I will give myself an out, by saying that the cost of energy may change my opinions in time. Gordon. PS there are so many variables, that comparing a windmill at one site is not objective to comparing with one at another site. I have trouble enough at the same site. I think that there is more value in being able to improve ones own windmill, rather than comparing to others. It is highly unlikely that even with a blueprint design that individuals would get 2 windmills that are the same. We normally only have a single unit we have made. There are many combinations of magnets,coils,air gaps,rotor dia,airfoils,loadings etc etc. I hope to improve my own recording of changes before I start with a new micro. become more energy aware |
||||
| dereksoftstuff Newbie Joined: 14/11/2010 Location: United KingdomPosts: 30 |
Gordon You make some good points. Thanks for your input. Infact increasing the envelope of wind speeds that the turbine can utilise well is the main aim, since turbine/generator/load matching is restrictive. This is probably where most of the extra power can be gained. The only way to compare windmill performance, is to have the data in the first place. So until everyone collects their data baselined to their anemometer, the present situation will continue. The commercial companies will continue to sell us a dream, and DIY'ers will post videos of windmills just spinning in the wind. The skystream does seem to have good feedback from their owners, some with real data logged over a year. I didn't know you could select different 'algorithms' via the user panel to check out other performance profiles. I will be experimenting with a buck configuration. The prediction is for dramatic energy price increases over the coming few years (just had 9% increase in UK). At a slight tangent - what are the typical power ranges of everyones windmills. I assume (jump in here) most are around 500W or lower, and only the really dedicated have > 1kW? |
||||
| VK4AYQ Guru Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Derick Unless I am missing the point here your statement of increasing the envelope of wind speeds captured as energy has a problem of collection in higher speeds, there is lots of energy so capturing it isn't a problem, and the problem is to control the speed of the mill, to do this with just electronics is fraught with danger should the system fail and the mill gets away, and destroys its self with over speed or burns out with over current, the simple furling system is the best of a bad bunch, and an active furling system, would be better but would suffer the same problems should its controlled fail, to me the blade furling / feathering system along the lines of a mechanical Governor would be the best way to go. Then the effort can then be concentrated on recovering the low speed energy, this is the case that is the most useful for increasing the usable output of the mill, Gordon has done a lot of work in this facet of energy recovery, primarily the use of voltage doubler caps to catch the bottom end energy. The problem is that there is very little kinetic energy in the wind at low speed to capture, so the simplest way is the best. I personally like the use of high voltage alternator with a down converter or the computer control you guys can do this will squeeze the bottom energy out without affecting the control of the mill. . The logical way to get low end energy increase is to use larger blades for more absorbed energy, and concentrate on the mechanical feathering of the blades as rpm increases. Just my sixpence worth All the best Bob Foolin Around |
||||
Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333 |
[quote]Just my sixpence worth [/quote] Bob, Put your horse and buggy away, and get the FJ out the shed, and spend your "20 cents" worth on a gallon of petrol. 
You are showing your age mate. 
Pete. Sometimes it just works |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Derek, I figured that you would opt for the hardest system to get right. The Skystream had about $US8,000,000 on development I believe. They opted for a boost converter front end. Good move there as well. The Skystream does an automatic self calibration [caled anemometer mode] to position the MPPT function. It does not need any other algorithms to select from. The R&D process obviously was rigorous enough to get the loading profile right without offering any options for the user. I hope that you realize that if the user gets the configuration wrong that it is possible to end up with fried components [particularly for a buck converter]. BTW my own windmills are presently a nom 500W and 1kW. I think others may rate them higher. I have recorded power levels higher, but this is the nature of windmills and logging. I have yet to fly a second 1kW mill. I eventually will have 2 x 1kW mills. My new windmills have approx 36cu in of N38 magnet volume, and this would relate to how much power it could realistically produce. I have some same dia N50 and 20% thicker, yet to test. During storms, the windmills may produce the measured 2kW peaks [what normally breaks electronics in the power train], but when you analyze the wind profile for the location, it seems pointless to concentrate on the storm performance, when it may happen only a fraction of the percent of the total time. Most likely is a power level less than 10% of the peaks [probably 99% of the time]. Windmill owners hope for a better wind distribution where they live, but reality comes in and we don't normally live where it is windy. If you have neighbours close by then efforts to MPPT may draw attention to your windmill with additional noise aspects. I hope you get a windmill and anemometer up to test in the real world. Once you do have a non MPPT windmill flying, then you will find how difficult it is getting correlation between the response of the windmill to the wind, as compared to the anemometer readings. I hope to be in a position to evaluate the TI micro soon enough. I don't have any piclog type data available on the net to view. Gordon. become more energy aware |
||||
| VK4AYQ Guru Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Pete I am working on the Chev Blitz at the moment, the FJ can stay in the shed for a while as it doesn't have a winch capable of pulling out my carrots. Can you get a gallon of petrol for 2/- down there. it might be worth coming down to buy a tankful. Horse now on a pension and wont pull the cart for me anymore, so have to use the blitz. 
All the best Bob Foolin Around |
||||
| Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333 |
[quote]Can you get a gallon of petrol for 2/- down there. it might be worth coming down to buy a tankful. [/quote] Theres a catch! You need the gallon tin. For 20 cents they will wave the nozzle past the fuel filler as long as the cap is still on.
You are doing it wrong, you are meant to feed the carrots to the horse. Bit OT. Pete. Sometimes it just works |
||||
| dereksoftstuff Newbie Joined: 14/11/2010 Location: United KingdomPosts: 30 |
Bob I suppose the way I look at it is like this. We have a choice of 2 bikes :- a) one fixed gear (e.g. penny farthing = pf) b) mountain bike (24 gears) We can choose whatever gear ratio we want for our pf, but once chosen that's it, it's fixed. Now it turns out that the wives have entered us into a bike race - the tour de france, which has some flat bits, some downhill bits and some uphill bits. Which bike do you want? Both bikes have brakes, and can be stopped (furled), which is independent of the gearing system. Gordon Early stages yet, and one never gives the user enough rope to hang themselves (validation etc). Or perhaps a big red button on the UI ... I was thinking of more like US$80 for development. Not all of the skystream users are happy! |
||||
| VK4AYQ Guru Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Derek Point taken, the brakes are the most important part of the equation, the only thing with your analogy of the bike is that you would be riding it and able to fix a problem, flat tyre or chain coming of then an d there but windmills spend 99 percent of their time on their own lonely and unattended with no one on the spot to fix what ever ails them. I await your your perfected design, as I am sure it can be done. All you need is a hundred times the electric knowledge that I have got, so go for it mate. With people like you and our learned members here anything is possible in the future. All the best Bob Foolin Around |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi readers, Derek has asked for some empirical data from our windmills. I have had some typical winds today. [In the 80% part of the 80:20 rule windspeed range] Here are some graphs of the logged data, and the performance curve and the actual recorded wind energy distribution. 
On the first graph, the pink line is the windspeed, the lt blue line is the windmill rpm, the red line is the power. The close to horizontal line is the battery voltage. The windmill rpm closely follows the windspeed. On the second chart, the green line is the expected output cubic power v windspeed. The red dots are the actual power v windspeed measured data. The third graph is the actual distribution of the wind energy, and shows the typical negative exponential type relation. I have had this translate into a typical binomial distribution during periods of sustained high winds. On this day was typical winds, and was the negative exponential. Peak power was 200W, peak windspeed was 13knots and peak rpm was 210rpm. The windmill has only series caps into the 24V battery and no dedicated buck MPPT. The caps are 3300uF 80VDC rated in the typical back to back arrangement. If I was to change the loading to 48V, I would need a cap doubler, but the results would be similar. I really can't complain about the power tracking achieved for my typical winds. These winds give 0.5-1.5kWh of power per day, from quite moderate winds. I know there is a place for a windmill MPPT, but maybe a hybrid of caps and electronics may offer a simpler solution. Gordon. become more energy aware |
||||
| dereksoftstuff Newbie Joined: 14/11/2010 Location: United KingdomPosts: 30 |
Gordon Good stuff - some real data! Your turbine rpm tracks the wind speed well - can you plot TSR over the same period in the chart (I assume that's about 1 day of logging). If you can, do you find it's fairly constant, and for your site (in general). The power curve data is also very consistent - does your chart plotter put numbers on the axes? I know this is beyond the pale, but do you know what difference not having the caps would make? I like looking at real data ... |
||||
| VK4AYQ Guru Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Gordon Looks very good, why fiddle with it? Hi Pete Found a picture of my computer tutor. 
All the best Bob Foolin Around |
||||
| Downwind Guru Joined: 09/09/2009 Location: AustraliaPosts: 2333 |
I just want to see how she cuts and pastes. I see why you had trouble connecting your pickaxe to the computer now. Next time use the one without the handle.
Pete. Sometimes it just works |
||||
| VK4AYQ Guru Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Pete Thanks for the hint, I was wondering why the wire with the ends on it was so long. I got a little pocket computer coming next week with Gordons assessment of the problem it has usb 2 ports on not 2 ports but 3 I think, it is new and kid proof they tell me but may not be Bob proof, so that may help, and no hard drive for me to stuff up you just plug a little card like a postage stamp with terminals on it to work and save things on, wish me luck. All the best Bob Foolin Around |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Derek, The windmill tested has quite thick airfoil section wooden blades [3 blades]. The TSR tracks at around 5. The TSR is quite constant across the windspeed range. Below cutin, the TSR approximately doubles, as expected [unloaded TSR]. The user sets maximum limits in a configuration page, and these are used for the graphing scaling on the plots. The analyzer is designed for change comparisons on an individual windmill. The important aspect is the green line. I have no need for numbered axis. Not having caps, results in a stalled windmill. There is some power, but a maximum of about 100W or so in a storm. The display of real time graphed data is very revealing on performance. The top graph above was for a 1 hour time period, recorded at 1 second intervals, from approx 10000 individual averaged time readings. This chart can be viewed continuously updated. I am hoping that there are other readers who can provide some alternative data as well. Gordon. become more energy aware |
||||
| GWatPE Senior Member Joined: 01/09/2006 Location: AustraliaPosts: 2127 |
Hi Derek, I have just rewired my windmill for 24V operation, and I needed to make a new cap doubler. Here is a graph showing the effect of the caps. 4700uF back to back in my normal cap doubler way. The lower dot graph is without caps. The green line is what we are trying to achieve. 
The matching of the lower end windmill output to the wind energy has been achieved without complicated electronics. It will be interesting to see how others do with their windmills once performance monitoring is achieved. Gordon. PS For my windmill, the bulk of the wind power that my windmill captures will be processed by the cap doubler. It is not until the power levels reach approx 20% that the main rectifiers play their part. The blend of the cap doubler with the main rectifiers offers a seamless process. The dot scatter is very tight, indicating no oscillatory behaviour. I cannot say the same for my testing with a micro. become more energy aware |
||||
| VK4AYQ Guru Joined: 02/12/2009 Location: AustraliaPosts: 2539 |
Hi Gordon It shows that simple is the best, the way it tracks the power available curve says it all, for a direct coupled windmill, high reliability the low tech is defiantly the way to go. All the best Bob Foolin Around |
||||
| dereksoftstuff Newbie Joined: 14/11/2010 Location: United KingdomPosts: 30 |
Gordon - Nice chart data comparison - looking good. Are there any other members who collect data, apart from Gordon and the picloggers? I'd be surprised if there aren't others who monitor their turbine's performance in one way or another. Even if you're using commercial units, the feedback would be welcome. Feel free to post any data or experiences you've had with your system ... |
||||
| Page 2 of 4 |
|||||
 Print this page |
