Hydro Generator on AC Output? ( A bit 'out there' sorry )

Modifications
, , , , ,
1

Hi folks,
this is a little out of left field but please bear with me…

I am considering options for an off grid system, for a friend…

We are talking about a 15KVA Quatro, BYD battery I think, I am not entirely sure, but I do know it is a stock implementation, currently with 5 kW of solar, AC coupled on the output, using a Fronius, a generator available on AC In 2 and a small hydro generator on AC In 1.

I am well aware of the 1:1 rule and reasonably confident that I have a handle on AC settings, weak AC, dynamic loading, boost factor… ETC

The issue is the Hydro, it is currently regulated by a governor, linked to a deflector, for speed and an AVR for voltage.
The overall regulation is OK but not great, especially when there is a step change in load. Obviously if it gets too fast or too slow AC In disconnects as it should.

The issue is the methodology… A deflector, by definition, is wasting water all the time, it has to so that it can both add and reduce generator torque, quickly. add to this the fact that the inverter is following the turbine generators voltage and frequency when it is connected and the whole thing starts to look, well; suspect, at best.

One option would be to rectify the turbines output and connect it via a PV inverter but that is a load of smoothing for a single phase generator and brings its own complications’, all be it without any of them being generator frequency.
This may be the best option, but I wanted to explore before committing to something I expect to be harder than I currently expect, to do well at least.

So that brings me to AC coupling…
In short, assuming it is properly synchronised, can I connect a generator to the AC output, exactly as a PV inverter would be, and leave the Quatro to manage the frequency by loading or unloading the generator?

I realise that in this mode the AVR would effectively need to be used to manage power factor, but that is a separate topic. (Save to say it will be required to change the voltage in response to reactive current so that voltage can drive the system I am talking about below.)

Given that the Quatro would be dominating the system, the turbine is only good for about 5.5kW, I would expect the system as a whole to operate like any other grid connected generator…
EXEPT…
I have no idea, and cant find much info about, specifically how the Quatro responds to voltage increase or how fast it can do that.

I would welcome any and all thoughts on the subject…
I have also listed below some of my assumptions and reasoning which may help guide comments without the need to ask further, although I am happy to answer anything.

Thanks for looking,
Al

Why…
Because I believe this would allow the Quatro to hold a stable frequency, forcing the turbine to adapt, or disconnect.

Testing and Caution…
I don’t want to hook this up and find that I missed something fundamental, only when a ball of molten copper and a load of smoke shoots me in the foot.

General…
I am very comfortable with AC connected motors in regeneration and grid connected alternators. I even have experience with C2C configured induction motors on single and split phase grid systems, as generators, some of them quite big.

I know what a synchroscope is and how to use/configure one, but on that note if connecting an alternator, or motor in regen, to AC out is a possibility, I wonder if I couldn’t leverage AC In to do the sync, as it normally would, and then connect AC OUT, to the generator, before disconnecting AC IN, allowing the Quatro to pull the frequency back to its setpoint.

On that thought, is there any way to connect to AC In, with its inherent protection, but then prevent the generator from driving the systems frequency, functioning as if it were an AC coupled energy source on the AC output. The two or physically linked at that point after all, so the differing behaviour has to be soft, right?

2

Given the power ratios for the Quatro and the Hydro, this may well work. You would need to be cautious about what happens when the batteries near full charge, as the Quatro may not be able to shut down the hydro power by raising the frequency to > 52Hz. If you can arrange the hydro control such that this condition is satisfied, then your only other concern is the PF balancing.
you will need to synchronise the Hydro to the inverter before connection, but you are already aware of that.
Unfortunately, this will only work on AC out, not AC in. Quatro feeding power to AC in is possible, but this relates to ESS type behavior, which is designed for grid coupled systems.

3

I am well aware of the frequency disconnect, or control, requirement but appreciate the heads up none the less as I may not have been.

I was going to say my plan, but its little more than musings right now, is to utilise Node Red to manage the turbine with a backstop of a frequency sensing, and possibly voltage sensing, relay to drop the connection should the code fail… This would be a lockout type function with a manual reset.

My hope is to have the inverter manage voltage and frequency, which ‘should’ keep the turbine output constant and more or less isolated from load swings, at least within the ramp rate capability of the charge/invert bridge, which I assume will be quite fast… I would like to know how fast but can’t find that info.

Obviously if there is no headroom in the batteries to act as a controlling load the hydro would not be connected or we would connect a dump load, perhaps a bulk water tank that would otherwise not be being heated. The latter would need to be art least as big as the turbines current output.

RE. Synchronisation…
Externally a synchroscope is an option, and likely the best one, I could build one but given the cost of them that seems like making work and taking risks I don’t need to.
However the Quatro has Sync built in and I had wondered about the following strategy…

Bring the turbine up to the current system frequency and voltage and get that stable, a small load being helpful in this respect.
Apply that signal to AC In, possibly via a current limit resistor…
This should cause the inverter to deal with the minor adjustments required to properly sync as opposed to being in the ballpark, V/F wise, and out of sync.
As soon as AC IN closes use the default, none critical loads, behaviour, or a flow that mimics it, to close AC Out 2, latch it and disable AC In 1
All of this would happen very quickly, long before the charger has started to ramp in and although this would effectively mean that AC Out 2 was linked to AC In 1, for a short period I believe that this is the case anyway when both are closed.
I realise that this may be true from an electrical perspective whilst the instrumentation and control strategy could well have a problem with it but I am assuming that worst case here is that AC In would disconnect, which is what I would want anyway.

From here it is the same weather an external sync or the strategy has been used.
There will be an external switch, whatever, because the FV monitor circuit demands it, and not using one would be very silly.

During and immediately after sync, the turbine would not be capable of delivering much, if any, power to the system and could well end up slightly motoring, which wouldn’t be a bad thing.

Once connected we would measure generator power factor, and adjust it by adjusting the AVR output setpoint, which, for a synchronous machine, an alternator in this case, would maintain an appropriate excitation level, relative to the system voltage, to match power flow to the torque from the prime mover. (At least that is how it works on 3 phase alternators…)

Once it is online… Add water, manage the power factor and it ‘should’ be stable

However there are lots of uncertainties there…
What I am hoping to find out is what the risks are and if I am missing something or making poor assumptions about the inverters control algorithm.

I am also well aware of the dangers of confirmation bias, so will not be clutching at the first reply that suggests I am not completely mad :innocent:

Please feel free to pull this apart and question whatever seems worth asking me about.

If anyone can shed light on the inner workings of the Quatro, at least as it pertains to this topic/question/s I would very much appreciate the opportunity to learn.

Am I correct in assuming that when AC IN is closed and AC Out 2 is also closed that they would be physically connected, internally, by nothing more complex than a couple of physical contacts?
If so I cant think of why connecting AC Out 2 to AC IN 1, for a brief period, would be a problem.

Anyone have an thoughts on that aspect of this topic?

Thanks for looking and commenting folks…

4

Sorry Mike, should have said…
Export, ESS controlled or otherwise, is disabled and would stay that way.

Thanks for your input.

EDIT…

The reason for asking about leveraging the Quatro to handle Sync is that sync relays/indicators can be had, used, from about £100 but at that price point would lack some functionality.
More capable units with properly settable parameters, offsets and stab time, are a little more expensive and by the time you get to a Deif GPC or the like, DSE coms to mind, you are into thousands…

The Quatro should, I think, be able to handle this naively, with a little creativity applied.

5

Usually when synchronising a new source to a system, the new source is brought to system frequency +0.1Hz, giving a 10 second period for synchronisation. Voltage is matched to within 5%. Most synchroscopes will need this. The contactor is then closed on the new source when phase matches within +/- 2 degrees. Load is then transferred by increasing the drive power to the new source.
Trying to switch the source between ACin2 and AC out would be at your own risk, this may cause problems with ground and neutral switching in the inverter. It may be that this also causes problems with frequency control, as the inverter switches from following AC in2 to being the frequency master. Given the relative costs of a synchroscope and the inverter, I would go for the synchroscope as the lower risk option - I used this to sysnchronise an unregulated 47kVA diesel gen to a larger inverter / battery system that only had one 3 phase AC connection, and no synch capability. The inverter control is fast enough to manage most load switching scenarios, up to its rated power. Variation in the Hydro power should be slowed to limit rapid swings, similarly to the permitted ramp rate for PV inverters.

Alternatively, if the hydro motor is a small synchronous motor, then start it dry on the quatro output, then add water…

6

That all sounds entirely sensible to me, and is exactly why I wouldn’t want to go making assumptions and risking the kit.
You are correct RE cost risk IMHO if risk cant be properly assessed when thinking about an entirely , lets face it; odd, proposal, then it shouldn’t be attempted.
I have no issue pushing boundaries but there has to be a solid understanding before going there.

I didn’t mention it above because I thought it would just complicate things and muddy the waters, no pun intended, but the Hydro currently has an alternator bolted to, well in, it so starting it as a load isn’t an option.
We could potentially re-engineer the thing and convert it to an induction motor.
A 230/400 3 phase motor, in C2C configuration can be operated using a DOL start strategy, with or without a current limit resistor, by getting the speed to a point slightly lower than synchronous, and then switching the mains and capacitors at the same time.
As the thing hasn’t got any poles until it is switched on, it will just motor at that point and typically isn’t much drama. They can also be conventionally synced.

I have Hydro sites I look after where this is done with motors up to about 80kW, onto split a phase mains supply, so 460V single phase as far as the motor/generator is concerned.
(And yes, I/we have those modified for the higher voltage)

I have also played with a configuration I call ‘Z’ which is essentially all three windings of a standard 3 phase machine in series. This yields twice the single winding voltage but limits the total current to that appropriate for a single winding. I cant remember what the derate is as a % but it isn’t relevant.
Think 12 wire machine in double delta as a single phase alternator, but with a few less windings in the mix __/
Neither C2C or Z will self start, but they both motor and generate well, with Z having a slightly better load profile , when motoring. (At least in my bench tests). Z also has a and less complicated capacitor configuration.

C2C has been around for some time, see ‘Nigel Smith, Sustainable Control Systems’ who responsible for making that a thing.
Z, I have never actually deployed as the risk, when its a customers site I am dealing with, wasn’t worth taking and I cant afford to fund even a moderate realistic test.

I am rambling, sorry…
The point is really, although changing the alternator for motor is possible it would be complicated because of the turbines layout. AND If I had a 3 phase motor I would just rectify its output, with a lumpy current waveform dumb diode bridge, and feed that, smoothed, to a PV inverter, as I have done many times in the past, all be it always to the grid in the past, but you get the picture I am sure.

I am not trying to penny pinch here, a physical conversion would be expensive so before we go there, or buy some form of active rectifier that makes rectifying single phase AC even doable, I wanted to explore any and all realistic options, and to do that I have to ask potentially silly questions.

Hopefully I get a consensus of answers and everyone gets a little entertained in the process…

WARNING…

I have in the past been accused of implying something was advisable when it wasn’t, or I was misinformed…
I am confident in my comments RE motor configurations BUT in isolation and or other circumstances, and NOT as they may or may not be potentially useful or damaging in this context.

To be absolutely clear I am NOT advocating any of this, and may never try any of it, depending on the consensus.

I claim no detailed knowledge of anything any Victron inverter is doing under the hood, beyond my understanding of its typical setpoints and ‘standard’ operation.

Cheers
Al

7

Actually, additional smoothing of the rectified output is not necessary, as this is built into the PV inverter…
As far as system integration goes, this may be the simplest solution…

8

I have spoken recently, withing the last 12 month, to several manufacturers of PV inverters, Fronius included and almost none of them would sanction using a PV inverter to handle rectified AC.

Ginlong , who produce Solis products, were willing to talk and commit to honouring warranty’s but even they wouldn’t or couldn’t give me figures relating to maximum permissible ripple, and insisted on a minimal amount of external smoothing.
(Not that I am convinced the figure was much better than a guess).

Why would you thing a PV inverter manufacturer would add expensive and bulky smoothing capacitors to a product designed to accept smooth DC from a PV panel?
Can you give me an example? It may be very helpful.

Are you aware of the current pulse issue, terrible power factor and high current spikes, when rectifying into smoothing caps? The more smoothing you need the worse it gets..

The Best solution to this sort of thing is a good SMPS, or two, that track AC voltage, or a slightly smoothed version of it, and boost that to the desired DC voltage.
Meanwell do a range of 3kW 250VDC supplies that are stackable, being isolated, making 500VDC @ 6kW relatively easily attainable.
I built a DC link system for a client, unifying a few small hydro systems, a couple of years ago using these supplies and PV inverters, which is working well.
I couldn’t find an inverter manufacture to play ball then either, we just had to roll the dice.

9

Of all the manufacturer’s I think only SMA has qualified versions of their inverter’s for Wind and Hydro generation. This seems to be significantly smaller market than Solar.
On other thoughts: You could use either an alternator or MDC motor as a hydro generator, and charge the batteries directly on DC…

10

SMA discontinued WindyBoy, the line you are referring to several years ago.
To be fair they could have brought out something I don’t now about, recently, but at last years green energy show in Birmingham they had nothing and didn’t want to talk about it at all.
For years SMA have taken the stance that they will not even sell you a product to use in any way other than a sanctioned one, let alone offer any backup.

I am not complaining about it it just is what it is!

There may be others but I have been using Ginlong products since ABB pulled out of the market and the products became PowerOne, which didn’t last long.
There are several small Hydro and wind companies doing the same, using Ginlong/Solis that is.

There 3 phase kit was quite big, at the time, I have sites where we installed multiple 30kW units many years ago, and I would be willing to bet they have gone larger now.
(Just checked, and yes utility scale if needs be but right down to 1.5kW at the low domestic scale)

There are also a couple of small hydro companies using random cheap imports, a strategy that will most likely bite them at some point I expect.

And yes compared to solar, wind and hydro, small scale stuff, are insignificant markets, and have been for the entire couple of decades, nearly, I have been involved with Hydro.

11

Sorry, stuff to do for a couple of days…

DC Charging may well be a good option, especially if the charger is current based as that would effectively decouple the hydro from the variable site load.

My primary concern there would be the charger itself, and what specifically to use.

There are several options. SMPS’s from Meanwell and the like that they would need custom integration, with one option there to be using the SMPS as an active rectifier and feeding its output to a PV inverter that we can properly integrate, and manage from Node-Red.

I guess its the typical dilemma… Cost and complication VS functionality and a robust design.

I don’t think too far out of the box is a good plan, unless there is a pretty large consensus that it will work and not risk damaging things, so the latter half of that equation is going to have to be the driver I think.

Solis PV Inviters are competitively priced, can be communicated with using Modbus RTU and I believe are compatible with Victron kit, using freq shift.
They can be easily controlled Grid: On/Off / Power Limit 0.1%-100%, whilst still having Frq control active, so a safe and usable option, all be it way more expensive than a direct connection, to the generator, via a contactor.

2X Meanwell CSP-3000-250 or CSP-3000-400 supplies would do the job, the former needing to be stacked to give about 450V Which is a supported mode.

Comments / Thoughts?

Al

12

Hi Al,
One thing to consider is smoothing the current waveform from a rectifier (3 phase or single) with an inductor, before further smoothing with a capacitor. This L/C smoothing used to be used a lot in the old days, but went out of fashion a bit with solid state.
Do you have some form of AVR / regulator on the hydro generation?
seems that there should be some regulation to prevent over-speed on light loads…
To my mind, hydro will work best supplying a constant, base load, and that Solar would be easiest to regulate quickly. So if the hydro is rectified and fed to the battery using a current limited dc/dc converter, and this power is less than the minimum guaranteed load, (or load plus available battery capacity to absorb charge) then essentially the hydro can run with limited regulation.