Multiple AC Source Synchronisation.

[NEO]

Dan,

    Is this the sort of circuit you had in mind.

Neither drawing or component electronics are my strongest attributes.

I found a voltage sensing relay that monitors voltage

Details

Range 1.1 to 26.5 volts

( 221Kb )

http://www.otherpower.com/images/scimages/7048/voltage_sensing.pdf

Would this work or should I say be suitable for, both to detect the low voltage at C1 and drive the contactor to conect the gen to the Busbar and if the voltage went up due to imbalance disconnect the contactor? It has a built in timer which I presume cause operation from time of fault detected between 100mS and 1S.......

As for my comfort range.. 3hz or less and within give or take + or minus 15 VAC of the output voltage.

Does that sound about right?

Thanks for your time in advance

Neo

btw what will it cost me for a circuit diagram for the inverter protection :-))

[robl]

Hi again Neo.  

On the older gensets droop was handled through adjusting variable resistances on the generator itself. They would generally be compound wound machines(with auxiliary windings) and I think the series winding would be the one that got adjusted. I expect Flux can explain this stuff better than I.

Droop is also very useful to avoid overloading generators that are intended to charge very large battery banks. If the current draw exceeds a preset limit the controller will actually let the voltage drop to avoid stalling the genny.

Synchronizing gensets is standard fare on most large ships. Until the 1980's these were mostly high voltage DC, not AC. As electronics advanced (esp. variable AC motors)it became cheaper to switch everything over. Despite the sophistication of the controllers, theres is always someone on watch, despite the installation of high-speed disconnects when blips occur. Anyone interested in this area should look for a copy of the Canadian Coast Guard Electrical code (or your country's equivalent) or troll Ebay or Amazon for books on marine electrical systems. The pictures alone are worth the cost of the books, but they generally include an envelope full of electrical diagrams inside the back cover.

On larger systems, alternators are replaced with synchronous generators and most of the problems go away, once sync speed is achieved. You will probably not find these units in the kind of wattages you want to use. However, you can construct a reasonable facsimili by using three-phase motors running as externally excited generators. You will need the grid or a pilot gennie to provide a sync source for the other units.

Anyway, good luck, and ponder my earlier suggestion to use multiple independent gen/load systems as being the cheapest/highest security route for remote power systems.

Regards

Rob

[maker of toys]

close, and it would probably work with the right resistor values and given your rather broad window of acceptable mis-match.  But I had something more on this order in mind:

everything else would be much the same.

note the voltage follower op-amp to avoid changing the RC timing of the circuit by sense loading.  if you choose R2 and R3 to be 1 megohms total, you can use the ordinary sort of 1/2W resistor mounted on perfboard.  going to lower value resistance makes the current and power disipation high enough to need wire-wound power resistors.  chose C1 as a high-voltage mylar type (2kV?) for reliablility. Additionaly, it may be advisable to add a couple kOhms of inrush limiting to stay below the current limit of the diodes on system start-up. . . maybe a hybrid of your circuit and mine?

(I'm assuming a 300 volt peak bus, which might be low if you're dealing with more than 208 nominal volts . . .)

Also, doing it as I have limits the effect of the diode drops on your sense voltage for better accuracy. your milage may vary. . .

 I think (based on absolutly no experience with the part in question) that the voltage sensing relay you mention will not leave you happy. . . . you want something bone simple so you can fix it with a hammer, solder and bailing wire when it quits in the middle of a blizzard.  in any case, the way i read the data sheet is that it would provide either operate on voltage condition  or relax on voltage condition, but not both, and further, getting the sense to work the way you want it to could be a bit tricky.

note that this circuit fragment is not a complete sense unit; it provides no self-protection, emi/rfi hardening or power factor sensing; it makes no attempt to sense malfunctions, nor does it attempt to load the gens fairly.  in other words, <include std_disclaimer.h>;  you will likely be money and aggravation ahead to buy a commercial unit . . . blak, blah, blah.

-Dan

(PS:  getting the design for inverter protection from me would be expensive and would have a LONG lead time; I have a bunch of other projects hanging fire including a couple of patents.  I understand Nando sometimes takes side jobs here. . . .but I don't know what his rates are.)

[NEO]

Hey Guys,

         With a caveat, I think I have found the solution to my problem or at least part of it in the form of this....

           File size    616Kb

http://www.otherpower.com/images/scimages/7048/Ump5bull.pdf"

Product #:   BE3-25A-C2D

Product Name:   Auto-Synchronizer (1 phase), Raise/Lower Contact Control Output, Raise/Lower Voltage Matching Contact Control Output, with Dead Bus

Description:   The industrial grade BE3-25A automatically synchronizes the oncoming generator to the bus. The BE3-25A can be selected to have summing point or contact outputs to control the governor or AVR to match the frequency, phase angle and voltage of the on-coming generator to the bus.

This unit features a Frequency Matching Raise/Lower Contact Output, Voltage Matching Raise/Lower Contact Output and Dead Bus.

List price is a 1000 dollars a unit.... which I don't think under the circumstances is not to bad....considering what's built in to the package,  less of course any  ancillary components/actuators.......... any one disagree?.....

For those in the know I would be grateful for an opinion. On this or perhaps an alternative....

Still need to source a protection device/circuit for the reference inverter any pointers in this direction would also be warmly received.

Thanks again

Neo

Ps Dan do you use the HAL system in your home ? if so how have you found it?

[robl]

Neo, you will need to budget for the governors as well...

Rob

[maker of toys]

HAL 9000 was in 2001: A Space Odyssey. He was, as you will recall, a difficult roommate. Just my pathetic attempt at an in joke.

I'm of the 'so simple you can fix it with a hammer' persuasion at home.

at work we interact with something like 40k discrete data points spread over 5 computer systems (and more than 22 PCs and macs) controlling several megawatts of powered equipment in a bad RFI/EMI environment; and we maintain 24/7/365 availability because the instruments are heavily subscribed. by the time I get home I don't want to be fixing the house, too. <G>

there is the odd touch switch and the like in the house, but nothing is both crucial and complex.

Most of my personal stuff is, well, 'toys' of one sort or another, (think analog electronic fuel injected, semi automatic spud gun) although my experimental mills and such are/will be well instrumented with ucontrollers and computers.

-Dan

[NEO]

Rob,

     Thanks for the reminder, the unit comes in 3 or so flavours and if I have interpreted the manual correctly the feed back control can be for various brands of Governor, I presume because of different comms protocols...

AMBAC international summing point governor

Barber- Coleman summing point governor

Woodward summing point governor

&

Raise/Lower contact output.

                   This was the flavour I was thinking of, correct me if I am wrong as I am thinking this is for servo mechanical control of the throttle/governor. Each engine has a mechanical droop governor. I would like to think I don't need to replace them, for one of the summing point types. I haven't heard back from the company to confirm this yet.

Also found this it's a start stop relay in one package, I initially thought about using the facility that is built into the inverters but thought this would be better.

I think it will handle all three of my generators too. It appears to be ideal, though I think I will need a timer on the stop side to be triggered to cover cool down.

         File Size 273 Kb

www.otherpower.com/images/scimages/7048/T2600.pdf"

Kind Regards

Neo

[robl]

Neo

That current relay looks to fit the bill. While you are at it, I would suggest you spend a little bit more money for a set of Murphy gauges, or equivalent. They are very common on marine engines. basically a dead-simple electro-mechanical system of gauges and sensors that give you "dead-man" over-rides in case of over-speed, under-speed, hi coolant temps, loss of cooolant and low oil-pressure. If you have ever been near a diesel that lost its governor you will appreciate the need for secondary shut downs...

Regards

Rob

[Colaman]

Skimming through all the comments, perhaps you would be better off with a single larger genset that is intermittently run to both supply sync'd AC and charge batteries in your system. And probably beef up the battery storage a bit.

In a config like this, the PV/Wind/Batteries will suffice for a certain number of days/hours, then the genset is operated for a period (eg, 8 hours) to top the system up.

Less components to source, anyway.

And for your info, perhaps you might look at the "Sunny Boy", "Windy Boy", "Sunny Island" series of inverters from SMA.

The sunny/windy boys are basically grid-tie inverters that take your PV or wind DC and convert to synced AC. In a standalone system, you make a local grid, with the sunny island as a master inverter-charger attached to a battery bank. It then controls a genset or input from an outside grid when needed if there's insufficient generating capacity on the local grid.

The sunny island will regulate power as necessary by fractionally shifting the grid frequency to regulate output from the sunny/windy boys to match system load and battery charging needs. If the system frequency is bumped up a little, the slave inverters will back off a bit and vice versa.

Basically that combo do all that you ask, but cost becomes the issue. The main advantage is that you can do it in increments. For me, I'd probably :

 - Buy sunny boy and PV array, link to outside grid and start saving cash immediately.

 - Buy sunny island and batteries, unhook sunny boy from outside grid, make a local grid and attach sunny boy to it, then link sunny island to both local grid to supply power and the outside grid to charge batteries when PV input is insufficient.

 - Buy genset and discard outside grid, or make it manually switched, or switched at off-peak rates if you've got them.

 - Buy windy boy and wind generators whenever cash permits it for supplemental power.

Another advantage of this is that someone else has done all this before and you don't have to dick around as much :-p

See http://www.sma-america.com/solar-technology/products/island-grids/battery-inverters/index.html

and a PDF that describes the system at

http://download.sma-america.com/smaprosa/dateien/1353/SI4500-11-BE2306.pdf

If you really want to wrap your head around it, go download the install manual for the SI-4500 from that site.

[mrpackethead]

Hi Neo

You say its easy to come by a G77 disconnect.  I googled G77 but did'nt find too much. Is G77 a standard of some sort or somethign else?

[mrpackethead]

Hi Neo

You say its easy to come by a G77 disconnect.  I googled G77 and found that G77 refered to a standard ( now superseeded by G83/1 ) that is relevent for systems smaller than 16A/phase.  

However what i could find was a 'device' that just did some kind of disconnect. Plenty of devices that are included as part of a entire system, but not just a disconnect device.  Can you point me in the right direction??

Regards

A.