Resistive heating controller advice needed

[mmurray70]

Hi guys, first post here. Looking to build a small turbine (2.5m-3m dia) for water heating and have some questions about a controller. I've done some reading and found people using everything from a fixed resistor directly connected to turbine, to home built boards and devices that switch in additional load, all the way to very expensive MPPT controllers. Just wondering what I should be looking at for my setup?

Can a fixed resistance work based on the blades running at a low TSR on low load, and running at a higher TSR at higher wind speed? This would certainly be the easiest option if its practical to use at all.

What about switching in additional load? I found a lot of discussion about this but not much in terms of proven examples. Many of the posts were very old too. What are my options now in 2019? Are there any products out there to buy these days to do this for a good price? Or good examples of circuits that I can build? Can I start with a fixed resistance connected direct and add an extra load with a SSR when the voltage reaches a certain point? Any suggestions?

[OperaHouse]

What size system? A dump load is generally associated with some effort to brake.  I don't think the decisions have gotten any easier.

[Adriaan Kragten]

The disadvantage of a fixed resistor is that the power curve is about a parabola. Measurements for a PM-generator loaded by different fixed resistors are given in my public report KD 78. The optimum P-n curve of a wind turbine is a cubic line. Assume that the resistance is chosen such that the point of intersection of both curves is obtained at a wind speed of 6 m/s. At higher wind speeds, the load will be too small and the rotor will turn at a higher tip speed ratio than the design tip speed ratio. At lower wind speeds, the load will be too large and the rotor will turn at a lower tip speed ratio than the design tip speed ratio. Below a certain wind speed, the generator torque will even be larger than the peak torque of the rotor and this means that the rotor will slow down to almost stand still.

So you need at least a device which makes that the load is disconnected from the generator below a certain rotational speed. The most elegant solution is that the resistance of the load is changed such that the optimum cubic line of the rotor is followed. This means that the resistance has to decrease at increasing rotational speed. But the heat capacity of the used resistors must be large enough to dissipate the maximum power of the wind turbine.

In stead of a resistance one can also use a battery charge controller which keeps the voltage constant. This gives a rather good matching for wind speeds above about 4 m/s and easy starting of the rotor as long as the open DC voltage is below the adjusted voltage of the voltage controller. A 27.6 V, 200 W battery charge controller is described in a free manual which can be copied from my website www.kdwindturbines.nl at the bottom of the menu KD-reports.

If you want to heat water with this device, the cooling plate has to be a part of the the water drum. Be alert that the contacts of the central transistor are not becoming wet. I am not sure if this device can have a maximum temperature of 100 °C. If the cooling plate is cooled by air, the maximum temperature will be lower. If you have to dissipate a lot of power, several 200 W dump load modules can be connected in parallel for one voltage controller.

[clockmanFRA]

I have found that Hugh Piggotts book,      .......  http://scoraigwind.co.uk/a-wind-turbine-recipe-book/   .....   has all his 40 years of Wind Turbine knowledge.   And Dan Bartmann & Dan Fink , s book Homebrew Wind Power goes a step further with details of a 17 footer.

He has sorted everything out, and gives length of blades and type, to match with load and output voltage required, the coil windings and counts, and explains types of controllers and many thing to do with a Wind Turbine installation.


With my 3 'babes' (3off 3.7m diameter/12 footers, Hughs design) I use 3off Morningstar Tristar PWM controllers that has settable battery voltages and charging rates, and once a voltage is reached or the batteries charged then each of the three Tristar controller diverts excess voltage to 2kW dump loads, large resistors.

The wind turbines are 'wild' even though they have auto tails/furling, that pull them out of the wind, even so a good battery bank will act as a good load for keeping the 'Babes' under some sort of control, and the controllers divert any excess.

I think some sort of controller will be required for your direct water heating, just direct connection to heater element will not allow the wind turbine to get going.

I feel a lot happier knowing that my 'Babes' are connected to a battery bank that can handle them.

[joestue]

Capacitors in series with the resistors can work. The capacitors resistance decreases as the frequency rises and the leading power factor increases the flux density in the generator.


Voltage controlled relays aso work. Will need maybe 6 resistors and relays

[JW]

link to  And Dan Bartmann & Dan Fink , s book Homebrew Wind Power goes a step further with details of a 17 footer.
 

https://www.otherpower.com/otherpower-store

[Bruce S]

mmurray70;
I'm curious about your TSR changes for low loads vs high wind.
Could you elaborate ?

Welcome aboard 

[SparWeb]

The trouble with a single resistor bank is that it's a constant load when the wind turbine isn't constant.
The resistors respond to load like a straight line.  The generator power curve is a parabola.  (and the wind is a cubic power curve, but that's a separate issue).
To see this you just have to draw a parabola and then a diagonal line through it... only intersects in one spot.

So if you match the resistor power to the turbine at some speed, then the resistors will be a heavy load at low speed, and a weak load at high speed.  Both are bad.
The load from the resistors when the turbine is barely starting to turn can prevent it from getting going... that's why many choose to switch the resistors off below cut-in speed.
The capacitors like Joe says can help bend that resistor power line into a curve, but still hard to match.
I've seen schemes to switch in additional resistors at higher speeds which can be done in DC or AC.
You can also fully rectify the AC output to DC and then regulate the power to the resistors with a PWM scheme.
All doable if you like to play with power electronics.  And if you don't, this is guaranteed to get you hooked.

[mmurray70]

Thanks to all who replied. Looks like I have lots of options and lots to learn. Seems like a fixed resistance is less then ideal, and the cost of a MPPT controller is mostly out of the question for now. So looks like I need to find a way to switch in additional load somehow. I would like to use a typical electric hot water heater with 2 elements. So lets say i'm running a 3 phase pmg, then rectify to DC to power elements. Would it be possible to start the turbine unloaded, then switch on one element when up to speed, and add on the other one when running at higher speed? I dont mind buying different elements to get whatever resistance I need. Would 2 stages be enough? I suppose I could weld in another port for 3rd element, or just have a 3rd load as an air heater. Just wondering how to switch them in at different points.

Capacitors in series with the resistors can work. The capacitors resistance decreases as the frequency rises and the leading power factor increases the flux density in the generator.

Voltage controlled relays aso work. Will need maybe 6 resistors and relays

Thats an interesting idea, first I heard of that one. So what would you use for capacitors? Just typical motor run capacitors? How would this be setup? Maybe a capacitor in series of each leg before going to rectifier? Any more details on voltage controlled relays?

mmurray70;
I'm curious about your TSR changes for low loads vs high wind.
Could you elaborate ?

Welcome aboard 

Just saying that the blades will run at slightly higher or lower then designed TSR depending on load. So because of this, it should sort of self regulate itself to match the load to a certain extent. See post 16 here: https://www.fieldlines.com/index.php?topic=149463.0  I "think" (not totally sure) this graph is showing a specific alternator/rotor running at slightly higher and lower TSR and the wind power curve is carefully placed in between if you get things matched just right.

[SparWeb]

Yes, that was what I was talking about when I made that graph.  It was just for demonstration, not a representation of anybody's particular machine. 
Just to re-hash what I wrote in the other thread:
The wind power curve has one shape, the generator power curve has a different shape.  They're never perfectly matched up but you can get them close enough.  So I plotted the graph above to show that one of the effects of the mis-match is that TSR is low at low wind speed, and it can get high for strong wind speeds.
Even with fancy controls on the resistance load, you still want a good furling system, otherwise the blades will run away in a storm.
And at the bottom end, you can see the power coming from the wind dropping way off, but the load from the generator is much higher.  That means the generator needs a few hundred watts to be turned at 50 RPM, but the wind power is only 10's of watts at 5 klicks.  You can't lift a 10 ton truck with a 1 ton jack.

So all that confirms what you seem to get, based on your comment about switching in additional load.  That works well to put an angle into the generator power curve.  You can make the generator power low when the wind is low, and higher when the wind is high.  If I was wiring up for resistance heating, I would have some beefy capacitors in line like Joe suggested, and tweak the furling to keep the top speed under control, then choose a low resistance so that the turbine isn't ever running fast.  Not until confident that it behaves itself under all conditions, and that the caps aren't going to blow up, would I then consider changing resistors or adding relays to switch loads on or off.  Just so that I could get a feel for how much is enough to add in, rather than making guesses, and accidentally letting it run away.  Remember, sizing the blades can be a bit of guesswork, too.

For the water heater, I can think of a few things that could go wrong with welding on another port, and it could be better to just get another (smaller) water heater in series with the first.  Two stages are much better than a single stage.  Like I said above, trying a very conservative single-stage to get a grip on the turbine's normal behaviour before complicating the controls would be the way I'd go.  Maybe you're comfortable with power circuits already, maybe not, it's just not safe to assume.  I don't even trust myself with this stuff.

To decide when to switch additional load on/off, you have two methods.  First, the voltage rises as the power rises, and second, so does the AC frequency.  You can use either one to signal the switch, or both for redundancy if you want.  A circuit to measure voltage can be tricked by spikes, and a circuit that measures frequency can be tricked by harmonics.  There are ways to deal with each of these but I believe measuring voltage is easiest (Joe can correct me if I'm wrong, there).  Then yes, just flip a relay when the threshold is crossed and suddenly 3 more resistors are on the circuit.

[clockmanFRA]

Maybe you're comfortable with power circuits already, maybe not, it's just not safe to assume.  I don't even trust myself with this stuff.

SparWeb is raising a very important point, especially in this day and age.

I meet a lot of electronic guys that make low voltage and high voltage electronics, and folk who programme things, but going into Power Electronics and High amperage areas is a proverbial mine field, and needs very considerate awareness.

My 'babes' run and direct connect to a clamping 48vdc battery, but they normally operate at about 56v to 61v, although the voltage seems low the actual amperage is severe and can be high, and therefore requires large, and I mean large diameter cables.

I have commented before that my 3off 2kW dump loads can get red hot, and the normal quadruple screw clamp connections just undo themselves and get hotter and hotter until failure.

 Not so much a problem with my three as the other 2 can take the strain and then one controller will say FUALT and no longer working and then I start inspecting things.
So Now I fully recommend using welded terminals and bolt together those terminals with a nut and bolt and a lock nut.

Now that I am older and wiser, although that may be subjective, !!!!!!!    If it was me, and I had the opportunity again I would wind my 'Babes' (wind turbines) for a mains voltage at say 140v to 350v and Grid tie them to my Mini Grid with appropriate controllers, as the cabling and connections are far more manageable.


   

[mmurray70]

A circuit to measure voltage can be tricked by spikes, and a circuit that measures frequency can be tricked by harmonics.  There are ways to deal with each of these but I believe measuring voltage is easiest (Joe can correct me if I'm wrong, there).  Then yes, just flip a relay when the threshold is crossed and suddenly 3 more resistors are on the circuit.

So lets say for example I use a few simple 3 or 4 pole ice cube relays on the AC side to switch in additional loads. I guess I could get different coil voltages to cut in at different points, or use a potentiometer in series with the coil to adjust when it trips. Seems simple enough. But what would you do to protect the first relay from the huge voltage range? Lets say for example first relay cuts in at 24v and you have peak voltages of 75 volts, that would likely cause the coil to burn out. What would you do here? Could a zener diode be used maybe to clamp the voltage to a safe range?

Would it be better to use solid state relays on the DC side? Seems like they have a wide range of input voltage (3-32v). Could probably make a voltage divider that would trigger them to start at different points and stay within the 3-32v range.

Also Im fairly comfortable working with high voltages and understand the hazards with them. I have more of a mechanical background but do have some basic electrical knowledge, but certainly no electrical engineer.

[SparWeb]

OK
The "ice cube" relays that I'm thinking of are automotive 12V only, so they'd fry after just a few cycles exposed to anything higher than 24V.  Plan to use gear with some hair on its chest.

The voltage trigger is a problem that's been solved many times before, and in many ways, so a little research will turn up lots of options rather than me trying to draw something.  It's not hard, but it takes some tuning to get it right, and there are many ways this cat can be skinned.  The potentiometer to adjust the trigger level is definitely a feature you can look for.  Zeners can clamp a small range of voltage by draining a small current - like a few milliAmps at a time.  Not a way to clamp 80 Volts down to 12V.  Voltage dividers, are simple enough at low current flow, so if you use it as the voltage being sensed, then you don't need a lot of current and it's not hard to figure out.

One thing you can expect is that your controller will need its own power supply regulated at a fixed voltage to run all of the brains.  An Arduino is one way to build the brains, but it won't accept just any power supply.  Most electronics are like that.  To get a steady power supply, expect to use a small battery, or a plug-in wall-wart of course.

[kitestrings]

In addition to the suggested resources above, you might try googling this board for information on your topic.  There are several folks over the years that have done, or attempted - we don't always know how they turned out - water heating.  One that comes to mind is a gent named Dave B.  IIRC he actually did a fairly large 15-18' dia. turbine dedicated to water heating.  You might even try sending him a PM.

There are also controllers out there that would allow you to divert varying amounts of power to heating water.  We do this with our MS Classic using a PWM auxiliary relay output to a pair of solid sate relays.  I suspect there are others, but my point is some people are adept at design, assembly, diagnostics, refinement of electronics - many of them frequent here - some not so much.  Trying to design a task dedicated wind turbine is one challenge, adding to that a control system for load matching/diverting, to me, would add another level of complexity that undoubtedly someone else has already worked through.

One last thought...I think of using the wind as a supplement when it comes to water heating.  For us it is water pre-heating.  There is likely never going to be a perfect match between the load, needing hot water, and the available power.  Maybe you have this in mind, but having a back-up fuel-source makes the outcome more likely to succeed.  We use a LPG, on-demand gas heater that modulates to accept pre-heated water.  The wind and PV at various times supplements the task very, very well.

Good luck.  ~ks

[mmurray70]

Ok the traditional relays are out. I went ahead and ordered some cheap 40a DC SSRs from ebay to do some testing with. The input voltage range being 3-32 volts gives a factor of 10 so I should easily be able to use a potentiometer as a voltage divider and stay safely within that range without worrying about any clamping circuits or anything. I might add a resistor from output side back to input to bump up the voltage slightly when turned on, creating a range between cut in and cut out.

Yes the purpose is water pre-heating with an additional water heater in a grid connected house to try and possibly save a few bucks on my bill some day. Ideally I would be better off with a grid tied system, but I dont want to mess with all the red tape and extra costs of doing grid tied properly. And realistically, Im only going to be able to generate a small amount of power, probably less then our hot water needs living with 3 girls.

An Arduino is another great idea, certainly give an awesome level of control. But Im going to try to stick with simpler controls first. And I want to stay totally isolated from the grid. Dont want to worry about turbine control going out if we loose power in a storm. A small 12v battery for control power could work if needed.

[kitestrings]

So, you're looking at swapping out elements to DC, correct, or maybe it will be a separate tank with DC elements?  And, maybe charging a small battery bank, or are you supplying the DC direct from the rectifier?  Those SS relays are DC - both the power switch and signal (3-32VDC).

My experience with these type of relays has been good provided that you:

over-size for the application
have adequate heat sink - they get hot, they fail
adhere to the recommendations for voltage suppression (MOVs &/or flyback diodes)

[JW]

I have used the SS switch being referred to here is a 40 amp DC mosfet, here is my P/N NTE.  RS3-1D40-41M

3.5-32vdc input    output 100vdc 40A.    to avoid transient current I put a 5KP100A diode parallel on output. 

[mmurray70]

Yes existing electric water heater tank will stay just as it is and im looking at adding a second tank to preheat the water, so less energy needed by existing water heater and still plenty of hot water when there is no wind to keep the girls happy. No battery bank, direct from rectifier to SS relays to water heater elements. I will change the elements to get the best resistance to match generator. I dont think AC or DC will matter as long as resistance is right. Im thinking I will actually be needing something fairly close to the original 3000w 240vac elements to match my generator, may not even have to change them at all. I ordered an industrial BLDC servo motor for a generator, waiting for it to get here so I can do some testing in the lathe and see what kind of numbers it puts out.

Will I need MOV's or diodes to protect the SS relays with heating elements? I thought that was mostly for inductive loads?

[kitestrings]

Okay, the preheat with a second stock tank makes sense, and keeps the household sane

The problem with the stock elements is not whether they are AC or DC as you've suggested.  Take a closer look at your example though.  A 3,000 watt, 240V element has a resistance of about 19 ohms, so say your turbine, via rectifier, is putting out 24VDC then you've only got 1.25A or about 30 watts.  I think you may need somthing more like these:

http://www.kansaswindpower.net/dc_accessories.htm
https://www.altestore.com/store/charge-controllers/dump-loads-dump-load-controllers/diversion-loadsdump-loads/water-heating-element-for-12v-or-24v-dc-1-npt-male-p2326/

You may want to even take another step back to say, how much will this thing produce?  What is the range of power I'm designing for.  A 2.5 m turbine in a 20 mph wind might do ~650 watts, and we'd do well to get all of what's available at the rotor, likely something less.  Going thru this logic though might help you frame the design parameters better.  I'd probably start with the turbine and adjust the load as needed.

With regard to the SS relays, I thought the same, but I think you will find that there is an inductive component to virtually any load.  If you look at the specs for most any Crydom, Teledyne, (we've been using VB Controls), etc. they all have recommendations for surge suppression.  The fixes are low cost and worth doing IMO.  Best, ~ks

[richhagen]

With my 60V solar dump I put a bit under 500 Watts into each 120V 2000 Watt water heater element, they are switched by a DC Solid State Relays through an Arduino which monitors the voltage.  It turns them on at 55 volts and the system can go as high as 60 when the solar overpowers my dump.  500 Watts is tolerable if you build your own tank and have lots of 1" (the size of common - read as cheap - water heater elements in the U.S. where I am) fittings in your parts box to weld into the tank. If the dump load fails, - and I have managed to burn up a couple of the buck 5V converters that power the Arduino controller- then for my solar, my controllers go open and stop the current into my storage bank, that would not work for wind though, so you need a more robust and redundant system. 

The main problem in your case, however, is that the power drops with the square of the voltage (from P=IV and V=IR, we can derive that P=V^2/R, so at 24V, a 120V element is only going to consume 24^2/120^2, or only 4% of the power it consumes at 120V nominal.  For 240V it is even worse, with 24^2/240^2 equal to .01 or 1% of the power consumption which is why, as Kitestrings pointed out, you get 30 Watts Consumption at 24V instead of the 3000 Watts at 240 for the 3000 Watt element. At 30 Watts per element, I think I might run out of those one inch fittings before I got a reasonable amount of power into my tank.  The 2000W 120V elements I found to be the best cheap ones for my system would only consume 80Watts at 24V and would do very little for a stock electric hot water tank with at most two element fittings.  There are 24V hot water elements available, but they cost many a multiple of the common cheap made for AC ones where I am at.   Rich

[mmurray70]

Im hoping to do much higher then 24VDC to save on wiring, elements etc. Doesnt make sense to use low voltage when not dealing with batteries. I didnt want to get into too much detail on the generator as I dont even have it yet. Bought the servo on ebay last week. Wont really know anything for sure until it gets here, but im thinking it has a voltage constant of 210v/1000rpm, resistance of 1.8 ohms and continuous current of 5.6a. These servo motors are rated very conservative, most people run much higher amperage then the cont rating, im hoping I can do the same.

I made a spreadsheet to calculate rotor rpm and power, given diameter, TSR and efficency. And also calculated current and power at load and loss in generator for a given rpm and load based on Sparwebs formulas in post 11 here: https://www.fieldlines.com/index.php/topic,149463.0.html

Again, the generator hasnt been tested so dont know anything for sure, but I made a graph using 2.6m rotor, TSR 7, 30% efficiency, one 19 ohm element switching on at 15kph and the second one switching on at 35kph. The total load on the shaft seems to follow the power in the wind fairly close over this range. I wonder how close blade calculations are to real world performance?

[Adriaan Kragten]

A method to calculate the rotor geometry is given in my public report KD 35. Chapter 4.3 describes the reasons why the real maximum Cp is lower than the Betz coefficient 16/27. A maximum Cp of about 0.45 is possible for rotors with twisted tapered blades and of about 0.4 is possible for rotors with constant chord blades and no twist. However, these values are only possible if the rotor is made accurately according to the theory and if the Reynolds value is high enough. This means that the product of the chord times the relative wind speed must be high enough. So a small rotor with slender blades needs rather high wind speeds to get sufficiently large Reynolds numbers. Long ago I have measured a 3-bladed constant chord rotor in the open wind tunnel of the University of Delft and the maximum Cp was a bit more than 0.4 at a tip speed ratio of 6. This rotor had a diameter of 1.8 m and wooden blades with a NACA 4412 airfoil and a constant chord of 0.083 m. The tunnel wind speed was about 10 m/s, so rather high.

The Cp is not the same as the aerodynamic efficiency. This has to do with the definition of the Cp. A Cp of 16/27 corresponds to an aerodynamic efficiency of 8/9. This is explained on page 15 of KD 35.

[SparWeb]

I wonder how close blade calculations are to real world performance?

Yeah, not perfect, but it is as accurate as you need it to be, especially at showing how it behaves as wind speed changes.  Understanding the behaviour is more important than getting the exact number right.  What you've done there makes sense.  Adriaan has the facts straight, but we could also drive home the point that you can do all the calculations you want, but with one pass of the spokeshave on your wooden blades the airfoil can change enough that it doesn't matter.  Any more math than that and you do it for the love of math (I do ).

The 19 ohm resistor looks like a good guess to start.  The choice of resistors actually depends on the resistance in the servo motor, so when you get your hands on that we'll know more for certain.  I don't quite believe the quoted "1.8 ohm" resistance yet.  I've measured the phase resistance of some similar sized motors and I would expect yours to be 3 or 4 ohms, but I had them connected in Star at the time.  What you measure depends on how it's connected, if it has connection options (Star/Delta) If the quoted phase resistance is in Delta, then I guess it matches.