Transformers to accompdate voltage drop

[PVfarmer]

But that is 2% of that circuit, not of the whole array. You do not add those 2% numbers to get the total losses for the system. If you have four strings that each lose 2%, your total losses are the same: 2%.

I'm honestly not aware of where I or the software did that.
I agree, if you have 8 DC combiner to inverter input "home runs", each loses 2% and all 8 lose 2%.

One thing- I goofed a bit- 4 AWG is 2.0% and 500kcmil is 1.8%, for the same 1000'.
Yet the 4 AWG supposedly puts out 0.2MWh more a year. That's odd.

I said-
"The software says 1.8% line losses using 4 AWG for that 16,000 feet. (Really 1000' distance)". (Also really 4AWG = 2.0%)

Each inverter has + and - to input A and + and - to input B coming from the combiner. So 4 conductors between each of the 4 inverter/combiner combos.
That's 8(+) of 1000' and 8(-) = 16 total @ 1000' = 16,000ft of 4 AWG. (And 1000' of EGC for each conduit used...)

I'm pretty sure it's calculating DC "line loss" for each "home run" at 2% individually *and* at 2% for all 8.

That's why I'm getting (in the pic I posted on page 2 of this thread):
16,000' of 4 AWG (8 "home runs") for 1000' distance/DC at 2.0% = 156 MWh/year
4,000' of 500kcmil (one "home run") for 1000' distance/AC at 1.8% = 155.8MWh.

 

[pcanning87]

I've worked on several multi-MW systems in NGRID territory. The transformer kVA rating (same as kW for our purposes) needs to be at a minimum the same as the AC system size. You can oversize the transformer if you like, but there's not really any reason unless you want to use a standard size.

I may have missed something in this thread but if your Point of Interconnection voltage is 15kV, then you need to step-up anyway. In this case I would definitely install a 1.0MW transformer at each array.

If you POI is at 480V then I would still likely install a step-up/step-down pair of transformers for the farther array. A 480V tie-in seems unlikely to me, I've almost never seen a system >1MW with a secondary Interconnection. In my experience, Systems this size almost always involve MV transformers, MV switchgear, reclosers, primary metering etc.


Sent from my iPhone using Tapatalk

 

[electrofelon]

I've almost never seen a system >1MW with a secondary Interconnection. In my experience, Systems this size almost always involve MV transformers, MV switchgear, reclosers, primary metering etc.

And my experience is the opposite. I have never seen a primary interconnection. I dont have a large resume of >1MW systems, worked on 3, toured another 3.

 

[PVfarmer]

1. The transformer kVA rating (same as kW for our purposes) needs to be at a minimum the same as the AC system size. You can oversize the transformer if you like, but there's not really any reason unless you want to use a standard size.

2. if your Point of Interconnection voltage is 15kV, then you need to step-up anyway. In this case I would definitely install a 1.0MW transformer at each array.

3. If you POI is at 480V then I would still likely install a step-up/step-down pair of transformers for the farther array. A 480V tie-in seems unlikely to me, I've almost never seen a system >1MW with a secondary Interconnection. In my experience, Systems this size almost always involve MV transformers, MV switchgear, reclosers, primary metering etc.

1. Sometimes you have to oversize it. For instance, if the OP was using 40 total 24kW inverters so 960kW, and they were SMA 24000TLs. SMA says the total inverter kW must be 90% max. of the step up (from 480/277V to MV) xfmr kVA rating. So a 1000kVA xfmr would be a no-go there.

2. Well...the inverters are 480/277V, so there has to be a step-up point. The question is where is it and who owns what. The OP/customer could get 2000kVA service at 12kV or 480/277V. With 12kV service the customer owns the step-up xfmr, and the losses or gains that go with it.
I personally see the prices of 1000kVA transformers and....look the other way!

3. I'm a little confused- there are no inverters that output 12kV, so again, there's always a step-up point- the OP is using 480/277V inverters.
The way I read their ESB, either A) NatGrid puts in the step-down from 12kV to 480/277V xfmrs, or B)the customer gets 12kV service and puts in step-up for the inverters.
MV transformers, MV switchgear, reclosers, primary metering etc.

With the B option, the customer is responsible for all that pricey stuff in italics, whereas option A is...simpler?
Pages 4 and 8 here:
https://www9.nationalgridus.com/niagaramohawk/non_html/constr_esb754759.pdf

 

[PVfarmer]

And my experience is the opposite. I have never seen a primary interconnection. I dont have a large resume of >1MW systems, worked on 3, toured another 3.

Well that's odd, both of y'all have NY in your locations!
But then, MA/RI/NH/NY all have different rules in the ESBs, so...

I still vote 1000', go DC!
I'm gonna have to think of a catchy slogan.

Did anybody go over those links jb posted? I agree with lots and disagree with parts of both.
In this one, for instance the italic part- that is not always the case!

In many of the systems we modeled, the overall wire losses were roughly evenly split between the source circuits and the output circuits, suggesting that the opportunities for conductor optimization are equal. This is not the case in reality, however. Even though the total length of the source circuits is many times longer than that of the output circuits, the incremental cost to upsize from 12 AWG to 10 AWG, as an example, is small compared to the incremental cost to upsize from 300-kcmil to 400-kcmil. Therefore, it is generally more cost effective to reduce wire losses by upsizing PV source-circuit conductors.
http://solarprofessional.com/articl...voltage-drop-conventions?v=disable_pagination

And then this one. Ouch my head. But same thing again- this is not always the case.
the overall wire losses were roughly evenly split between the source circuits and the output circuits
If you change meaning of source circuits to be the DC side and output circuits to be the AC side, lower amps/much more footage of conductor/much smaller AWG seems to win for 1000'.

Increasing the conductor size to reduce voltage drop increases its cost (since cost is proportional to cross-sectional area) and decreases its energy losses (which in the denominator are inversely proportional to cross-sectional area). The result is no net gain, and we would have the added expense for the larger cable if we did so.
http://solarprofessional.com/articles/design-installation/optimal-dc-cable-selection-in-pv-designs?v=disable_paginationhhh

 

[electrofelon]

1. Sometimes you have to oversize it. For instance, if the OP was using 40 total 24kW inverters so 960kW, and they were SMA 24000TLs. SMA says the total inverter kW must be 90% max. of the step up (from 480/277V to MV) xfmr kVA rating. So a 1000kVA xfmr would be a no-go there.

Its odd SMA makes that 90% statement. Seems like one of those silly manufacturer statements that doesnt make any sense. Besides, if the utility owns the transformer, we have no control over how they size it.

 

[PVfarmer]

The issue is getting the power from the 1.44MW array back to the utility transformer which is around 1000' from the interconnection point.

Disregarding the voltage drop question. What are the main parameters for sizing a utility transformer for a solar site of roughly this size?

Its odd SMA makes that 90% statement. Seems like one of those silly manufacturer statements that doesnt make any sense. Besides, if the utility owns the transformer, we have no control over how they size it.

Um...
I think I get it- the OP does have to pay for the transformer!
I don't get the SMA/90% thing either, but it doesn't seem like the POCO should make the customer void a warranty (if it would do that?)

So people are saying the OP would be ok with a 1000kVA xfmr for their 40 x 25kW = 1000kW of inverters.
But the OP is asking about *one* xfmr in their 2nd quote.

Since the larger array is closer to PCC, and will need >1000kVA, maybe... one 2500kVA for both arrays?
Then you would have to use DC for the 1000' run!

I had a feeling LA would differ from MA- kinda looks like customer pays here. I just skimmed it...

http://www.entergy-louisiana.com/global/documents/utility/conn_small_elec_generators.pdf

 

[electrofelon]

Um...
I think I get it- the OP does have to pay for the transformer!

Well its kinda semantics: Regardless of where the service point is and who owns the transformer, the customer is paying for it one way or another - unless the utility is subsidizing the cost of the project. I did an installation (non solar) where the utility specifically stated that their tariff is such that they dont subsidize the MV installation costs against future electric usage or any other reason. So its kind of interesting: If the client cancelled service, could we keep the $13,000 500kva transformer? I doubt it. Potentially an interesting legal question for someone with nothing better to do!

 

[PVfarmer]

Regardless of where the service point is and who owns the transformer, the customer is paying for it one way or another -

If the client cancelled service, could we keep the $13,000 500kva transformer? I doubt it. Potentially an interesting legal question for someone with nothing better to do!

You kind of lost me in the 1st part.
The OP for instance.
The "grid voltage" is 12kV, inverters are 480/277V.
So... what's the metering voltage, 12kV or 480/277V? If the tariff for PV output is metered at 12kV, then the customer sort of "owns" the xfmr losses.
If POCO did put in the xfmr and PV was metered at 480V, the customer wouldn't own the losses.

I'm thinking a 12kV to 480/277V xfmr of 2500kVA size isn't a standard thing (in places?), so the customer/OP is paying so that the POCO doesn't have to worry about removing it if in fact the service is cancelled...whereas a xfmr the POCO has in stock can be put back in stock if cancelled?

Funny, 2500kVA happens to be the largest NatGrid lists in the chart of those they "procure"... pages 99-100:
https://www9.nationalgridus.com/non_html/shared_constr_esb750.pdf

9.3 AVAILABLE FAULT CURRENT

For equipment rating purposes, the following tables list the maximum fault currents available at the Company’s transformer secondary terminals. These fault currents are based on the lowest impedance of transformers the Company procures and on infinite supply impedance on the primary side.

 

[GoldDigger]

And of course, by infinite supply impedance they really mean infinite supply admittance!
Infinite bus means essentially zero source impedance.

 

[electrofelon]

You kind of lost me in the 1st part.

IF the transformer is on the load side of the service point, the customer pays for it. IF the transformer is on the utility side of the service point. the utility buys it and charges the customer for it. Customer pays for it either way just saying. Regarding losses, sure, who pays for the losses would be determined by where the metering is.

 

[PVfarmer]

IF the transformer is on the load side of the service point, the customer pays for it. IF the transformer is on the utility side of the service point. the utility buys it and charges the customer for it. Customer pays for it either way just saying. Regarding losses, sure, who pays for the losses would be determined by where the metering is.

I'm reading it as they furnish/procure up to 2500kVA... don't those both mean "pay for"?

1.2.1
Provisions
The installation provisions and costs shall be in accordance with the Company's filed tariffs in the applicable State.
Company furnishes, owns...
MA&RI
•an outdoor padmounted transformer
•primary cable

3.1
Service Voltage and Maximum Transformer Size:
Wires 4
Delivery Voltage 480 wye/277
Transformer kVA Limit Outdoor
Padmount 2500
Vault (3- 1 ph. transf.) 1500
Vault (padmount) 2500

https://www9.nationalgridus.com/niagaramohawk/non_html/constr_esb754759.pdf

 

[electrofelon]

I'm reading it as they furnish/procure up to 2500kVA... don't those both mean "pay for"?

I appear to not be getting my point across. You build a 1 MW PV plant. The utility is going to extend their line and supply a bunch of MV equipment for free? No, they are going to charge you, and the cost of the transformer is in there somewhere.

 

[PVfarmer]

I appear to not be getting my point across. You build a 1 MW PV plant. The utility is going to extend their line and supply a bunch of MV equipment for free? No, they are going to charge you, and the cost of the transformer is in there somewhere.

I just don't see how the Company can charge the Customer for something the Company owns. After all, they're a POCO, not a bank!
The customer has to sign an easement that allows the company equipment on customer property...

I really think it depends where you are.
The POCO in Louisiana gets zero RECs- I think that's why they don't pay for xfmrs.
NatGrid gets RECs and those RECs are often worth more than what they pay out in tariff, so they pay for standard equipment.

1.0 INTRODUCTION
1.1 Purpose
This Supplement to Electric System Bulletin (ESB) #750 provides specific guidance for Customers whose service point is the secondary side of a Company owned transformer located on Customer owned or controlled property.

 

[electrofelon]

I just don't see how the Company can charge the Customer for something the Company owns.

As I said in post #28, it is an interesting quandary. In those cases, Maybe its yours when you are done with it, not sure never tried.

 

[PVfarmer]

As I said in post #28, it is an interesting quandary. In those cases, Maybe its yours when you are done with it, not sure never tried.

Anyhoo, we're getting off-topic here.

Can we agree on these?
1. From what we know, the OP is probably going to want one 2500kVA xfmr for the ~2250kVA or so of PV.
2. That xfmr should go as close to the MV grid/PCC/connection point as possible. (Keeping in mind that the 480V side of the xfmr could in fact BE the PCC...)
3. So back to part of the original OP question... I'll say it again- the best way to avoid voltage drop over a 1000' run, in my opinion, is going DC for the 1000'.

Interesting thing- when Helioscope does the DC line losses, it uses STC amps *and* 480V, for 1000VDC.
So when it tells you 3.2% losses with 6 AWG, that's a pretty big over-estimation...whereas for AC line losses, it uses 480V for 480V.