Ideal xfmr(s) for connecting 480/277Y inverters (60kW total) to 12470/7200 grid

Status
Not open for further replies.

GoldDigger

Moderator
Staff member
Location
Placerville, CA, USA
Occupation
Retired PV System Designer
Any grid tie inverter (except hybrid inverters) will match the voltage seen coming from the grid and just supply current at that voltage. It also used a number of specific tests of frequency and voltage stability and very low source/sink impedance if the grid. If the grid source is not seen the GTI cannot, by design and regulation, produce any output.
When a qualified grid source is seen the GTI has no choice but to dump all available PV power to the "grid", not knowing how much of the net power is going to local loads and how much is going out through the meter.
 
this doesn't help

this doesn't help

If the grid source is not seen the GTI cannot, by design and regulation, produce any output.
When a qualified grid source is seen the GTI has no choice but to dump all available PV power to the "grid", not knowing how much of the net power is going to local loads and how much is going out through the meter.

um...doesn't page 4 contradict itself there? :roll:

Page 18-
Integrated Plant Control
The inverter can display the Q(V) characteristic curve specified by the grid operator by means of
Integrated Plant Control without measuring on the grid-connection point. The inverter can
automatically compensate equipment installed between the inverter and the grid-connection point
after having activated the function (for information on the system configuration refer to the Technical
Information "Integrated Plant Control and Q on Demand 24/7" at www.SMA-Solar.com).
http://files.sma.de/dl/21561/STP12-24TL-US-IA-en-16W.pdf

Page 4-
The inverter can automatically compensate equipment installed between
the inverter and the grid-connection point.
(next sentence...)
The function "Integrated Plant Control" is not capable of compensating irregular or fluctuating reactive power demands
? e.g. from connected machinery ? when the machinery is connected between the inverters and the grid-connection point.
http://files.sma.de/dl/7418/IPC_STP-TI-en-13W.pdf
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
Right. But!
First of all, what's going on here?
It's in German, that doesn't help.
It looks like the SI is only hooked up to ONE Of the 3 phases of the grid here. Huh?
http://www.sma-america.com/fileadmin/_migrated/pics/summenstrom_01.jpg

And then, never mind that- regarding your quote above-
If the MDP and grid are both getting power from the Tripowers, and the MDP includes the SIs, and then the grid goes down, the Tripowers will throttle down and supply the load instead of grid, won't they?

No. The Tripowers are utility interactive inverters and will shut down if the grid goes down.

The SI has two AC connections. AC1 is the critical loads connection and will continue to be powered by the inverter from the batteries if they have a charge. AC2 is for the grid (or another generator) and automatically disconnects if grid shuts down.

Thus if the Tripowers are connected (ultimately) to AC2 on the SIs (through a transformer or whatever) they will shut down if the grid goes down. I have no idea if the Tripowers can in any way be configured to connect to the critical loads panel (AC1) of an SI setup. It would almost certainly make more sense to use single-phase SMA inverters instead.

This goes with that link/pic above.
Although the Sunny Island is only connected to one line conductor, it automatically manages the total power. The energy meters work in a cumulative manner, so that the power from the individual line conductors is added up. The Sunny Island can then always feed in or draw as much energy as is needed in accordance with the sum total of all phases
. http://www.sma-america.com/products...ed-self-consumption.html#sthash.N6XDLcQN.dpuf

Like Golddigger I am skeptical or confused. It's curious that they wrote that page to have an animation, but it's not actually there, and both of the links below that section are dead. One hopes its not something that a marketing guy made up for their new website that doesn't actually accord with the technical specifications. Note also that the language you quoted does not refer to Tripowers.
 

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
um...doesn't page 4 contradict itself there? :roll:

Page 18-
Integrated Plant Control
The inverter can display the Q(V) characteristic curve specified by the grid operator by means of
Integrated Plant Control without measuring on the grid-connection point. The inverter can
automatically compensate equipment installed between the inverter and the grid-connection point
after having activated the function (for information on the system configuration refer to the Technical
Information "Integrated Plant Control and Q on Demand 24/7" at www.SMA-Solar.com).
http://files.sma.de/dl/21561/STP12-24TL-US-IA-en-16W.pdf

Page 4-
The inverter can automatically compensate equipment installed between
the inverter and the grid-connection point.
(next sentence...)
The function "Integrated Plant Control" is not capable of compensating irregular or fluctuating reactive power demands
? e.g. from connected machinery ? when the machinery is connected between the inverters and the grid-connection point.
http://files.sma.de/dl/7418/IPC_STP-TI-en-13W.pdf


Power factor and reactive power have nothing to do with the functionalities we've been talking about so far. That's a whole other discussion which is not relevant. And rather than a contradiction, what you quoted is basically saying that these inverters have some capacity to deal with reactive power demands, just not a whole lot.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
Ok- so in this "not line-side" diagram, it would go from SBs into the SIs and/or out to grid.
It would go in that setup at a point after the customer owned PV meter and before the MDP.
http://www.nationalgridus.com/non_html/dg_sample_1line_a.pdf

But then in this "line-side" diagram, the SIs would go...into the main breaker (busbar?) of the MDP, right? By that I mean "the single breaker that the other 4 breakers are attached to".
Because, the 480/277 from the SBs is going through a 480/277 to 120/240 xfmr on its way there.
So- in this second scenario, no "loss of grid" for the SBs, and no "PV connection" for the SIs?
https://www.nationalgridus.com/non_html/DG_sample_1line_B.pdf

Will the SBs not "throttle back" without SIs to "tell them to" or something? How would they even know the "grid went down" other than the output going from unlimited to being whatever the load is at that moment? If they're already taking care of the load AND the extra is going out to the grid...
Isn't the issue more that SBs don't interact with batteries than that they won't put out power when the grid is down? Like you said, there's a way to do it?

P.S.


I don't know enough to be the first one, so I'm just confused! ;)
In an SI/SB system, the AC elements are, left to right, a Sunny Boy, a protected loads panel, the Sunny Island, the MDP, the service, and the grid. The only connection between the SB and the grid is through the SI, and when the grid goes down everything to the right of the SI is de-energized. The SI inverter turns on and begins supplying power to the protected loads panel and providing the qualifying AC to the SB, but its integrated transfer switch disconnects this microgrid from the MDP. It has to by the laws of man and physics. If your PV inverter(s) is/are connected directly to the MDP, they will shut down when the grid goes down; they cannot "see" the AC being produced by the Sunny Island.

Since the connection between the PV inverter and the MDP is through the Sunny Island, the size of the PV inverter (its max AC current) is limited by the size of the OCPD that connects the SI to the MDP. In the case of a Tripower you would first have to transform its output voltage (phase to neutral) from 277V to 120V and connect it to the MDP through three Sunny Islands, and if you are using 5kW SI's you are limited to a single 15kW Tripower.

Any scenario wherein the PV inverter is connected directly to the MDP will result in it shutting down when the grid goes down.
 
No. The Tripowers are utility interactive inverters and will shut down if the grid goes down.

If you had 3 SIs connected, one on each phase of a single Tripower, that single Tripower would NOT shut down, would it?
I guess I could run 3 of the 4 15000 TLs straight into the grid and run the 4th one through the existing electrical system.
As in, 4th Tripower --> 3 SIs/batteries --> MDP?

I have no idea if the Tripowers can in any way be configured to connect to the critical loads panel (AC1) of an SI setup. It would almost certainly make more sense to use single-phase SMA inverters instead.

Like Golddigger I am skeptical or confused. It's curious that they wrote that page to have an animation, but it's not actually there,

Yep, that bugs me too, I'm going to ask them.
Um....this here? It seems to have something to do with what we both said?
http://www.aeesolar.com/sites/default/files/graph-ac-coupling.jpg

That's from this article.
http://www.aeesolar.com/grid-tied-solar-systems-with-backup-power


And rather than a contradiction, what you quoted is basically saying that these inverters have some capacity to deal with reactive power demands, just not a whole lot.

Yes, I get that now. "Equipment" and "machinery" are different things there.
 
In an SI/SB system, the AC elements are, left to right, a Sunny Boy, a protected loads panel, the Sunny Island, the MDP, the service, and the grid. The only connection between the SB and the grid is through the SI, and when the grid goes down everything to the right of the SI is de-energized....and if you are using 5kW SI's you are limited to a single 15kW Tripower.

Any scenario wherein the PV inverter is connected directly to the MDP will result in it shutting down when the grid goes down.

15kW is more than enough for times of no grid.
Is there a catch re: putting 3 of the TPs into the grid and the 4th TP into SIs/batteries/MDP/also into grid?


The batteryless grid-tie inverter will shut down at the start of a blackout, but will turn back on (after a mandatory 5-minute waiting period) when AC power from the battery inverter is detected and supply AC power to the backup subpanel, and if enough power is available, will be used by the battery inverter to charge the batteries.
http://www.aeesolar.com/grid-tied-solar-systems-with-backup-power

That seems to make sense.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
15kW is more than enough for times of no grid.
Is there a catch re: putting 3 of the TPs into the grid and the 4th TP into SIs/batteries/MDP/also into grid?
As I mentioned, you will have to transform the Tripower's output down from 480/277V to 208/120V to work with the Sunny Islands.

You'll have to dance with the AHJ as to how the interconnection will look. You mentioned wanting to interconnect at medium voltage, but unless you have primary metering (metering at MV) or some sort of feed in tariff available to you, you'll have to transform the other three Tripowers' output to whatever your service is. If your service isn't three phase you may have to abandon the idea of Tripower inverters completely.
 
well, that's weird

well, that's weird

Module DC Nameplate80.0 kW
Inverter AC Nameplate 68.1 kW
Load Ratio: 1.17
Annual Production113.3 MWh
Performance Ratio 85.4%
kWh/kWp1,416.2

That's with two 24000TLs and 1 20000TL, 240 panels into the 24000s and 80 panels into the 20000.
The 24000s grid connected at 480/277, and the 20000 going down to 208/120 and then back up to 480/277.
Ok.

Now I'm totally baffled, the annual production went UP 2 MWh. Hmm.
 
As I mentioned, you will have to transform the Tripower's output down from 480/277V to 208/120V to work with the Sunny Islands.

That doesn't seem to be too much of a problem, output wise.

you'll have to transform the other three Tripowers' output to whatever your service is. If your service isn't three phase...

That's what I'm trying to figure out how to do.
The service is 12470Y/7200, so three legs of 7200V and a neutral.
I would think if the POCO gave us 3 xfmrs, 30 kVA each total of 90kVA, going from 7200/480, then the 3 wires of the Tripowers, which are combined at the AC panel, would go into 1 xfmr each?
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
That doesn't seem to be too much of a problem, output wise.



That's what I'm trying to figure out how to do.
The service is 12470Y/7200, so three legs of 7200V and a neutral.
I would think if the POCO gave us 3 xfmrs, 30 kVA each total of 90kVA, going from 7200/480, then the 3 wires of the Tripowers, which are combined at the AC panel, would go into 1 xfmr each?

You keep talking about separate LV/MV transformers for the Tripowers; why wouldn't you combine them first and transform the sum? Combining at MV is a non-trivial exercise.

Also, you say that the service is 12470Y/7200; are you being metered at MV? Usually, unless you have a special arrangement from the utility, you must interconnect behind the meter.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
Module DC Nameplate80.0 kW
Inverter AC Nameplate 68.1 kW
Load Ratio: 1.17
Annual Production113.3 MWh
Performance Ratio 85.4%
kWh/kWp1,416.2

That's with two 24000TLs and 1 20000TL, 240 panels into the 24000s and 80 panels into the 20000.
The 24000s grid connected at 480/277, and the 20000 going down to 208/120 and then back up to 480/277.

There is a maximum current that you can push through a Sunny Island when the grid is up. I have concerns that the output current of a 20kW Tripower transformed down to 208/120V will be too high to go through the SI's. Also, do we know that the Tripower has the capability of throttling back its output in response to frequency munging like a Sunny Boy will?
 
Last edited:

jaggedben

Senior Member
Location
Northern California
Occupation
Solar and Energy Storage Installer
That's what I'm trying to figure out how to do.
The service is 12470Y/7200, so three legs of 7200V and a neutral.

That's not the service. That's the utility's distribution feeder. Try to keep terms straight. The existing service, as you told us in the first post, is 240/120 split phase. You are considering asking them to change your service to 3 phase 480/277 because it suits your desire to use Tripowers. (Why do you want to use those? I'm still not exactly sure.)

There might be several other options, such as having them upsize the 240V service, or getting two services, etc. It all depends on what you're trying to achieve, what your state laws allow regarding net-metering or feed-in-tarrifs, and whether you're willing to consider other options for inverters.

I would think if the POCO gave us 3 xfmrs, 30 kVA each total of 90kVA, going from 7200/480, then the 3 wires of the Tripowers, which are combined at the AC panel, would go into 1 xfmr each?

ggunn already gave a good response to this. But in addition, you have it wrong. If you have three transformers from the POCO (i.e each with three windings) then you'll have no combining at a panel. This sounds to me like it would actually be 3 services and I have no idea why they would offer you that. If you have a combiner panel, it should go to a single 3-phase circuit that would be your service. What transformer(s) they use to bring that to the service point pretty much doesn't need to be your concern.

Perhaps this should have been said a while ago, but for someone working on a relatively large project you seem not to understand some of the basic concepts. Or, at the very least, you're not proof-reading your posts as much as you should be. :huh:
 
From what?

As opposed to all 4 Tripowers going straight into the grid- I took one of them and put it into the "load area" via the MDP, and the output went up vs. 4 straight in.
I guess what that means is two 24000TLs into the grid and a 20000TL into the load are a bit better than 4 15000s with the same 320 panels total.
IBM should get that Watson computer on this thing, I think...


You keep talking about separate LV/MV transformers for the Tripowers; why wouldn't you combine them first and transform the sum? Combining at MV is a non-trivial exercise.

Also, you say that the service is 12470Y/7200; are you being metered at MV? Usually, unless you have a special arrangement from the utility, you must interconnect behind the meter.

The service now is one 7200V to 120/240 on a pole, which goes overhead to the meter and MDP in a shed outside of the building.
Then there are actually two...subpanels (house and barn) off of that.
Like ben said, I was mixing up terms.
The distribution feeder is 12470Y/7200.
Existing service is one of those 7200V wires into a 25kVA 120/240V can on the pole.
Potential upgrade is all three 7200V wires into XXX xfmr for a 480/277 service, 100 amps.

The 100 amp service could come from either a 12.47 to 480/277 3 phase xfmr OR three 7200V xfmrs wired in Y to 480/277.
So, those three 7200V wired in Y would be "combined at MV" to the Tripowers...how? Directly?


Also, do we know that the Tripower has the capability of throttling back its output in response to frequency munging like a Sunny Boy will?

It seems to say that it does in the manual, but I'm going to check with SMA.
 

ggunn

PE (Electrical), NABCEP certified
Location
Austin, TX, USA
Occupation
Electrical Engineer - Photovoltaic Systems
As opposed to all 4 Tripowers going straight into the grid- I took one of them and put it into the "load area" via the MDP, and the output went up vs. 4 straight in.
I guess what that means is two 24000TLs into the grid and a 20000TL into the load are a bit better than 4 15000s with the same 320 panels total.
IBM should get that Watson computer on this thing, I think...

What software are you using to run your simulations?

The service now is one 7200V to 120/240 on a pole, which goes overhead to the meter and MDP in a shed outside of the building.
Then there are actually two...subpanels (house and barn) off of that.
Like ben said, I was mixing up terms.
The distribution feeder is 12470Y/7200.
Existing service is one of those 7200V wires into a 25kVA 120/240V can on the pole.
Potential upgrade is all three 7200V wires into XXX xfmr for a 480/277 service, 100 amps.

Your service is not 12470/7200V, it is 240V single phase, and you are probably metered at 240V. You cannot connect a Tripower to 240V single phase. You cannot interconnect outside the meter without a special arrangement with the POCO. IMO what you need to do is replace your service with a 480/277V service and a transformer to service your loads, interconnect your Tripowers at 480/277V, and forget about MV.

The 100 amp service could come from either a 12.47 to 480/277 3 phase xfmr OR three 7200V xfmrs wired in Y to 480/277.
So, those three 7200V wired in Y would be "combined at MV" to the Tripowers...how? Directly?

I would combine the outputs of the Tripowers at 480/277V and step up from there with a single transformer, assuming that interconnecting at MV is possible.

About the ability of a Tripower to regulate its output according to frequency:
It seems to say that it does in the manual, but I'm going to check with SMA.
It either says it or it doesn't, but by all means call SMA if you aren't sure.
 
Last edited:
ggunn already gave a good response to this. But in addition, you have it wrong. If you have three transformers from the POCO (i.e each with three windings) then you'll have no combining at a panel. This sounds to me like it would actually be 3 services and I have no idea why they would offer you that. If you have a combiner panel, it should go to a single 3-phase circuit that would be your service. What transformer(s) they use to bring that to the service point pretty much doesn't need to be your concern.

Perhaps this should have been said a while ago, but for someone working on a relatively large project you seem not to understand some of the basic concepts. Or, at the very least, you're not proof-reading your posts as much as you should be. :huh:

I mean 3 single phase 7200V to 480V transformers there.
As far as the Tripowers, they seem to have the best features, and are a lot more flexible in short.

Yes, I am totally clueless- I'm not the electrician here, or the POCO. The electrician (Master's license) would prefer if I deal with the POCO.
Basically, again, I want to be able to suggest something to the POCO fellow (who is quite nice) that may be less $$ or better some other way.

This is all about feed-in tariffs, there are all kinds of procedures and complications- 60 kW happens to be one of those points where it's "less complicated".

Three single phase xfmrs wired in Y would connect to a 480Y/277 MDP...how?
Or if line side, to the inverters...directly?
Is there a problem having the 480Y from the inverters go BOTH out through the junction box AND in to the load through a 120/240V xfmr (paid by customer)?

The output of the PV is a lot more than any loads, 90% of the time. So it seems it would make sense to feed the PV into the load at 120/240, then the "unused" part goes straight into grid.

I guess I got the 3 single phase idea here on this forum!
" from what I understand on such a distribution system most of your pole top type transformer banks would usually be connected wye on the primary side - mostly because they build the banks out of commonly stocked single phase transformers that are connected line to neutral on the primary side."
3rd comment in this thread.
http://forums.mikeholt.com/showthread.php?t=161313
 
What software are you using to run your simulations?

Your service is not 12470/7200V, it is 240V single phase, and you are probably metered at 240V. You cannot connect a Tripower to 240V single phase. You cannot interconnect outside the meter without a special arrangement with the POCO. IMO what you need to do is replace your service with a 480/277V service and a transformer to service your loads, interconnect your Tripowers at 480/277V, and forget about MV.

I would combine the outputs of the Tripowers at 480/277V and step up from there with a single transformer, assuming that interconnecting at MV is possible.
About the ability of a Tripower to regulate its output according to frequency:
It either says it or it doesn't, but by all means call SMA if you aren't sure.

The software is SMA Web Design (it's free).
The existing service is technically called single phase three wire 120/240.
Like this:
http://www.cruisersforum.com/forums/attachment.php?attachmentid=92481&d=1417198738

Sorry, what's MV? Just Mains Voltage?
 
I would combine the outputs of the Tripowers at 480/277V and step up from there with a single transformer, assuming that interconnecting at MV is possible.

These are the options. The 2nd to last is best, right?
This should be just readable size wise.
However, the service is 200A, more than twice the output of the PV.
I guess 208/120 4 wire 200A would work too, but that seems wrong when the majority of power is going to be 480/277 from the PV.

ServiceCapture.jpg
 
Status
Not open for further replies.
Top