tips on diversion load controller without batteries(so no charge controller)

[junglehydro]

Hi all, first post here! 
 
So I am at the point of designing my system at the moment, and have come to the point of a diversion load controller. I will not be using batteries, so I do not have a charge controller. Instead I am looking for some recommendations on a standalone diversion load controller.

My situation:

I live off grid in the jungle of Colombia, where we are currently powered by two 320watt panels and some $#|+ty gel batteries(24v system) that I put together. quality stuff can be hard to come by here, but at least my charge controller/inverter combo has been proven reliable for the past 6 months.

As we are building an off grid eco hostel on our property we're going to need a lot more power in the future, and my neighbor has a large creek that runs year around. I have done some measurements and we are probably looking at a head of 100-120 meters over a distance of 300-400 meter of pipe. (3")
I am looking at installing a 5kw turbine at the bottom and running a 240v power line to my property (about 500m).
Since the 5kw will be our peak usage, I don't want to install any batteries and instead run directly off the turbine. But for this reason I am looking at a separate diversion load controller that I will hook up to 5x 1kw water heating elements.
Any advice greatly appreciated!

[SparWeb]

Hi there
Welcome to Fieldlines!

Just going by the first impression, I wonder if you will have power quality and stability issues.  A couple of issues, or dumb questions, take your pick:
What makes your turbine run at exactly 240V and 50/60Hz (if you don't have an inverter)? 
(if you do have an inverter) What feeds your inverter the exact DC mix it asks for?
How do water flow fluctuations affect the turbine output power quality?
How to changes in electrical load (turn on the electric oven) affect the turbine?

You need some inertia in your system.  The term is a bit abstract for an electrical system but the analogy is suitable and it is a serious concern.  People usually rely on batteries to stabilize their power supply, though you are right, they don't have to.  But if it's not batteries, it has to be something else that is resistant to fluctuations (the grid, flywheels).  The inverter receiving the input power has to have the supply available when it needs it. 

Maybe you intend to control the feedstock water flow.  You might be feeding the turbine a steady flow of water, but the flow that keeps it at 1800 RPM when the load is <1kW is a lot less than the water flow when the load goes to >4kW.  What kind of actuator is going to respond in time before the inverter kicks off-line?

I may have extrapolated too far with the information I have here.  Please tell us more; this sounds interesting in any case.

[clockmanFRA]

Hydro.

Hugh Piggotts site has some nice real working Hydros.

He's a Wind turbine man normally, but been installing Hydros in Europe for many years.

He also has lots of info on control systems.

http://scoraigwind.co.uk/

[junglehydro]

Thanks for the welcome!

And thanks for the link clockmanFRA, I found his site before but maybe need to take another look :-)

What makes your turbine run at exactly 240V and 50/60Hz (if you don't have an inverter)? 

  A hopefully constant flow of water

(if you do have an inverter) What feeds your inverter the exact DC mix it asks for?

  So not using an inverter, it comes out at 240 volts.

How do water flow fluctuations affect the turbine output power quality?

there shouldn't be any as I am building a setup that should have constant flow year round.

How to changes in electrical load (turn on the electric oven) affect the turbine?

So I am looking at building a load controller, that can measure load and divert all extra load to a system of water heating elements.

So the plan is to have the turbine run constantly, produce the same wattage.

> measure the output of the turbine > measure the demand from my side > deduct the demand from the amount produced and send the surplus to a diversion load water heating system. So the demand as far as the turbine goes is always constant, and the load controller takes care of the fluctuations.

The only thing is that there's not really an off the shelf load controller that does this as far as I've found, so looking at putting something together with an arduino controller.

[bigrockcandymountain]

Sounds fairly possible to me.  I think i would do 5 dump controllers in parallel, one for each element.  That keeps the amps way down.  They can be off one arduino though i think.  There are some difficulties with pwm controlling of AC power, and I'm not an electronics guy.  It also depends on the generator end.  Is it permament magnet or does it have a wound rotor and automatic voltage regulator? I think your dump load will kind of become your voltage regulator so if it has an avr, you may need to bypass it.

I would say set the controller to keep the ac volts between 230v and 250v.  Starting large motors like pumps etc might be a problem.  What will you be trying to run off of it?

Hopefully i talked enough nonsense that the adults will come and teach us both what you should do..

[Scruff]

I made a few designs for DC diversion "not involving batteries".
I haven't implemented any because they're expensive and clunky.

The working principle was to use a token battery that would be always full. ~10Ah to act as a voltage clamp and signal to the DC Diversion Charge Controllers.

[SparWeb]

No nonsense.  Let's see what JH comes back and says about the loads in the system. 
Well pump, kitchen stove... big stuff like that.
I'm still concerned that there would be brown-outs and bright bulbs as the voltage swings, in response to loads switching on and off.
Some people have figured out how to operate MPPT grid-tie inverters without the grid.  I don't know how.

[junglehydro]

  I think i would do 5 dump controllers in parallel, one for each element. 

Yeah that's what I am looking at now. I have a guy helping me out that knows electronics and can program arduinos and we're designing a system like that. So if anyone here has experience with that and can warn me about pitfalls I am eager to know.

Is it permament magnet or does it have a wound rotor and automatic voltage regulator? I think your dump load will kind of become your voltage regulator so if it has an avr, you may need to bypass it.

Yeah it's a wound rotor with regulator, good one on the bypass. I hadn't thought of that.

Hopefully i talked enough nonsense that the adults will come and teach us both what you should do..

That's what I go for mostly as well 

No nonsense.  Let's see what JH comes back and says about the loads in the system. 
Well pump, kitchen stove... big stuff like that.

The bigger stuff would be like 1800 watt powertools and such., but if the controller responds fast enough with the dump load this shouldn't be a problem right? And I can live with some bulbs briefly dimming when a heavy load is switched on, I don't nee to forget that I live on a remote jungle location 

The working principle was to use a token battery that would be always full. ~10Ah to act as a voltage clamp and signal to the DC Diversion Charge Controllers.

Yeah that would probably work but indeed a bit clunky :-)

[mab]

I have a 300w chinese PMG water turbine that is rated 220v / 50Hz, designed to run on 12-14m head. It came with a regulator that is basically a heating element and a triac with a voltage sense cct - probably a bidirectional zenner/tvs diode & resistor. difficult to say exactly what it is as it's potted. Probably senses the peak voltage going past 350v-ish and turns on the triac. It's fairly crude, but it works fast and is simple, and keeps the peak voltage below the point where things start blowing up.

as i'm running 18m dynamic head and a long transmission line, and i've got batteries at the other end i don't use the regulator normally, just let it run 300+ volts.

Point is I suspect that a voltage sensing diversion control will be easier and have a faster response to load disconnection than trying to measure load currents. Something like the above might be relatively easy to cobble together and get you up and running whilst you find a more elegant solution.

[OperaHouse]

This is what I use as a dump to my water heater.  Makes a nice shunt controller.

[Bruce S]

Operhouse;
I was hoping you would post in this thread. Your programming classes for the Uno's might also be used in this scenario.

Cheers
Bruce S

[junglehydro]

This is what I use as a dump to my water heater.  Makes a nice shunt controller.
(Attachment Link)

Don't really know enough about electronics to know what I'm looking at here...

[OperaHouse]

When things slow down  I'll do a thread on it. This is my third 40 gallon water heater in the garage used only for clothes washing. I feed hot water into the washers cold inlet so all cycles use hot water. Cloths come out steaming and cleaner than at home.  This is all excess PV power not used by the charge controllers.  The washer also works off solar panels directly, no battery.

[junglehydro]

That is so cool! So that would be very much like what I want to do, create a full hot water system, in my case with a hot water line running to showers and a little pool. But I like this idea a lot so might implement the cold water lines from the washing mashine as well. I'm assuming you do regulate temperature so you don't end up boiling your clothes? We are now looking at a temperature controlled solenoid valve that adds cold water whenever a temperature threshold is reached.